SEARCH

SEARCH BY CITATION

Keywords:

  • body size;
  • egg size;
  • Geometridae;
  • Lepidoptera;
  • life history;
  • Ornstein-Uhlenbeck;
  • phylogenetic comparative methods

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

We present a phylogenetic comparative study assessing the evolutionary determinants of egg size in the moth family Geometridae. These moths were found to show a strong negative allometric relationship between egg size and maternal body size. Using recently developed comparative methods based on an Ornstein-Uhlenbeck process, we show that maternal body size explains over half the variation in egg size. However, other determinants are less clear: ecological factors, previously hypothesized to affect egg size, were not found to have a considerable influence in the Geometridae. The limited role of such third factors suggests a direct causal link between egg size and body size rather than an indirect correlation mediated by some ecological factors. Notably, no large geometrid species lay small eggs. This pattern suggests that maternal body size poses a physical constraint on egg size, but within these limits, there appears to be a rather invariable selection for larger eggs.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Progeny size is a trait of central importance for the theory of life history evolution. For example, larger newborns, eggs, and seeds have in numerous cases been reported to show higher fitness (Braby, 1994; Williams, 1994; Fox & Czesak, 2000; Moran & Emlet, 2001; Ficetola & de Bernardi, 2009). However, producing larger progeny is more costly for the mother; therefore, progeny size is central to the quality–quantity trade-off that mothers have to make (Bernardo, 1996; Fox & Czesak, 2000). Currently though, despite a long history of studying this trade-off (Marshall, 1953; Salthe, 1969; Berrigan, 1991; Olsen & Cockburn, 1993; García-Barros, 1994; Kinnison et al., 2001; Czesak & Fox, 2003; Forbes et al., 2010), both ultimate and proximate determinants of progeny size are far from being fully understood.

Across various taxa, there exists a general pattern that larger species tend to give birth to larger offspring. At the interspecies level, correlated evolution of these traits has been demonstrated in a phylogenetic context in hymenopterans (Blackburn, 1991), some crustaceans (Poulin & Hamilton, 1997), fish (Crespi & Teo, 2002; Cooke et al., 2006), frogs (Byrne et al., 2003) and birds (Dyke & Kaiser, 2010). In several other taxa, a correlation between these traits has been studied but without explicitly considering phylogeny (ladybirds –Dixon & Hemptinne, 2001; whiptail lizards –Schall, 1978; turtles –Van Buskirk & Crowder, 1994; rotifers –Walz et al., 1995; birds –Rahn et al., 1975).

For Lepidoptera in particular, despite the demonstration of allometric scaling between mother and offspring size in some groups (García-Barros & Munguira, 1997; García-Barros, 2000, 2002), quantitative and causal aspects of this relationship are still unclear. In terms of causality, the allometric scaling could either (i) be based on a purely mechanistic (anatomical) constraint (maternal body must be large enough to fit large eggs), (ii) a developmental constraint resulting from ontogenetic integration (genes for large body affect egg size as well) or (iii) be a consequence of ecologically based parallel selection on both egg and body size, with no causal links between the two. Maternal body size as an anatomical constraint on maximal egg size is obvious in insects such as tsetse flies, which lay a single offspring at a time, of comparable size to the mother herself (Buxton, 1955). However, the situation is less clear in species that are able to carry tens to hundreds of mature eggs simultaneously, like Lepidoptera (Wiklund & Karlsson, 1984; Wiklund et al., 1987; Fox & Czesak, 2000; Bauerfeind & Fischer, 2008) and the majority of other insects (Hinton, 1981). In turn, a developmental constraint would imply that the two traits have some genetic link. Indeed, genetic correlation between offspring size and body size has been reported in other taxa (e.g. birds –Larson & Forslund, 1992; Diptera –Avezedo et al., 1997), including Lepidoptera (Fischer et al., 2002). However, the existence of a rigid developmental link between egg and body size is inconsistent with the frequent observation that egg size is not or is just weakly affected by environmentally based reductions in body size (Tammaru et al., 2002; Bauerfeind & Fischer, 2008). This suggests that there is enough leeway for variation in egg-size body-size ratios (Wiklund & Karlsson, 1984).

In addition to constraint-based explanations, a positive association between egg and maternal body size could have an adaptive basis. Such a relationship may be selected for indirectly through correlation of optimal values of both egg and maternal body size with the values of some third trait, perhaps an ecological one. Indeed, it is reasonable to assume that optimal progeny size, and consequently the optimal balance between progeny size and number, will vary across different environments (Fox & Czesak, 2000). An intuitive assumption is that larger eggs give rise to more viable progeny (Smith & Fretwell, 1974), and various hypotheses citing ecological advantages of larger egg size have been established across various taxa (e.g. Palmer, 1985; Honěk, 1992; Fox, 1994; Carriere & Roff, 1995). Consistent with this basic assumption, Lepidopteran species with larger offspring (from larger eggs) have been shown to be able to feed on a wider range of host plants, including those less favourable (Wasserman & Mitter, 1978; Nakasuji, 1987; Braby, 1994). Additionally, in polyphagous species where females are less discriminating about laying their eggs on particular host plants, it is advantageous for offspring to be able to disperse further afield (Tammaru et al., 1995). Larger offspring have also been shown to disperse further (Wellington, 1965; Barbosa & Capinera, 1978). Furthermore, for offspring that feed on tougher woody plants as opposed to softer herbaceous plants, it has been shown that those feeding on the former come from larger eggs (Reavey, 1992). Also, host plants are considered to be more numerous and more easily found in forest habitats compared to nonforest environments, which necessarily bears consequences for various life history traits (Barbosa et al., 1989; Snäll et al., 2007). Environmental factors that could determine egg size are not necessarily linked to diet or habitat. For example, it is also hypothesized that larger eggs are a survival mechanism in species that overwinter in this stage (Reavey, 1992).

Previous studies in selected groups of Lepidoptera, which consider egg size and maternal body size in a phylogenetic context (e.g. García-Barros, 2000, 2002), have been limited to the commonly applied comparative method of phylogenetic independent contrasts (Felsenstein, 1985). Such analyses cannot go beyond testing correlated evolution. In this study, we use a recently developed method of phylogenetic comparative analysis based on an Ornstein-Uhlenbeck model of trait evolution, which enables us to examine potential adaptive relationships of traits, involving directed movement (although there is also a stochastic component) towards specific optimal values. These optima can themselves evolve on the phylogeny and can be modelled on categorical factors or continuous measurements of the selective environment (Hansen, 1997; Butler & King, 2004; Hansen et al., 2008; Labra et al., 2009).

The primary aim of this study is to understand the factors that underlie variation in egg size in geometrid moths, using an original 53-species data set. We chose the Geometridae as our study group, because they display large variation in maternal body size, egg size and ecological habit. The life history of geometrid moths is well documented and is now starting to be understood in a phylogenetic context (Snäll et al., 2007; Wahlberg et al., 2010), although egg size has never been systematically studied in an evolutionary ecology framework. Here we test the hypotheses that maternal body size variation is a potential determinant of egg size variation, along with four external ecological factors. These ecological factors are tested to see whether they have effect on egg size either directly or possibly indirectly (via influencing maternal body size). These factors are larval host range (poly- vs. oligophagous species), tree- vs. herb feeding in larvae, habitat preference (forest vs. nonforest species) and overwintering stage (eggs or not).

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Study organisms

The Geometridae are a well-characterized, distinct monophyletic group of Lepidoptera (Minet & Scoble, 1999; Hausmann, 2001, 2004). Nevertheless, there is considerable variation in life history traits (e.g. Barbosa et al., 1989; Snäll et al., 2007; Javoišet al., 2011) including body and egg sizes (Brehm & Fiedler, 2004), as well as in habitat choice, dietary habit (polyphagous or oligo/monophagous, tree or herb diet) and overwintering stage (Koch, 1984; Leraut, 2009). Such a high ecological diversity makes this group very attractive for studying correlates and potential determinants of egg size (or other life history traits) in a phylogenetic comparative framework.

Phylogeny

For constructing a phylogenetic hypothesis (Fig. 1), we principally used DNA sequence data for Geometridae provided by Snäll et al. (2007), Viidalepp et al. (2007), Õunap et al. (2008), Wahlberg et al. (2010) and Õunap et al. (2011), but also a small portion of unpublished original sequences. The final data matrix comprised 116 species and 4643 positions, and the tree was constructed in BEAST (Drummond & Rambaut, 2007) using Bayesian inference. As results of comparative phylogenetic analyses are more easily interpreted if ultrametric trees are used (Hansen, 1997; Butler & King, 2004; Hansen et al., 2008), we constructed an ultrametric tree using BEAST 1.5.4 (Drummond & Rambaut, 2007). The 116 species tree was constructed as part of some ongoing research into geometrid phylogeny at the University of Tartu. For the purposes of the present analysis, the tree was pruned to leave just the 53 relevant species for which we had life history data. Further details of the phylogenetic analysis can be found in Appendix S1.

image

Figure 1.  Ultrametricized phylogeny of the 53 species of geometrid moths used in this study. Relative (%) and absolute (time) scale bars indicating tree height are shown below. Sample sizes and life history trait values are provided for all species.

Download figure to PowerPoint

Life history data

To obtain data on maternal body and egg sizes, the moths were reared in the laboratory on their preferred host plants. Larvae were either collected from the field, or adult females were collected with their offspring being reared in the laboratory. A day after adult eclosion, each upreared female was killed by freezing, air-dried in the laboratory (23 °C) and weighed. The abdomens were then soaked in water and dissected. Lengths and widths of three mature (chorionated) eggs were recorded for each individual.

Sample sizes of female individuals per species ranged from 1 to 48 (Fig. 1) depending on how many specimens could be reared. Dry weight of the freshly eclosed adults served as the measure of body size in the analyses. Using insects reared in standard laboratory conditions and of fixed adult age minimizes environmentally induced variation in egg and body size (García-Barros, 1992; Oberhauser, 1997; Fischer et al., 2003). Egg size was calculated as the volume of a prolate ellipsoid based on measurements of egg length and width as consistent with egg shape in geometrid moths (Johnson et al., 1987). For each individual, the mean value of three measured eggs served as the estimate of egg size.

Data for habitat and larval diet were obtained from Leraut (2009), and data for overwintering stage were taken from Koch (1984). All ecological variables were coded in a binary form. Species were designated as being either oligophagous or polyphagous, and also as either tree or herb feeders. We considered oligophagy in a strict sense to indicate that a particular species feeds on just two specific genera of plants at most. Habitat preference was split into forest and nonforest dwelling species. Values for life history traits are given in Fig. 1.

Phylogenetic comparative analyses

To explore the determinants of egg size, we performed the comparative analysis using the ‘SLOUCH’ package (Hansen et al., 2008; Labra et al., 2009) written for the R statistical environment (R Development Core Team, 2005). SLOUCH performs phylogenetic comparative analyses based on an Ornstein-Uhlenbeck (OU) model of evolution in a maximum likelihood framework. The OU-based method enables us to model the evolution of a trait across a phylogenetic tree in response to different selective regimes (Hansen, 1997; Butler & King, 2004; Hansen et al., 2008). These selective regimes themselves can be values or states of other traits, or those of the environment (predictor variables). SLOUCH works with continuous response traits whereas predictor variables can be both continuous and discrete (Hansen et al., 2008).

When studying a regression relationship between two continuous traits – egg size and maternal body size in the present case – it must be decided which one of these traits should be treated as a response or dependent variable, and which one as the predictor or independent variable. To do this, we examined the strength of the phylogenetic signal in each of the traits taken separately: indeed, it makes sense that a trait exhibiting a weak phylogenetic signal would be more free to vary in response to changes in the state of the other trait (Hansen et al., 2008; Labra et al., 2009). The strength of the phylogenetic signal is captured by a metric termed phylogenetic half-life [when the SLOUCH regression is performed with only an intercept (i.e. no predictor variables)], being expressed relative to the height of the tree (Labra et al., 2009). Accordingly, following Hansen et al. (2008) and Labra et al. (2009), the trait exhibiting a smaller half-life is identified as the response variable. The underlying logic is as follows: (i) if one trait evolves quickly, producing an essentially random pattern with respect to phylogenetic relationships and the other is strongly influenced by phylogenetic proximity, an observed strong correlation between the traits cannot be due to a shared phylogeny; (ii) this leaves us with the possibilities that the observed correlation is generated by either (a) both traits responding to a hidden third variable, (b) the traits responding to each other or (c) one trait being dependent on changes in the other; (iii) it is however highly unlikely that a trait responding to changes in another, randomly evolving trait with respect to the phylogeny, would generate a strong phylogenetic signal in the responding trait by chance, and (iv) therefore, if there is a causal relationship between the variables, it is more likely that the trait that is more free to vary with respect to the phylogeny is responding to changes in the phylogenetically constrained trait. Naturally, causality cannot be rigorously proved in a macroevolutionary framework and there is no way to discount the influence of all possible ‘hidden third variables’, but these methods provide a means to make educated inferences regarding the directionality of possible causal relationships.

When a trait exhibits phylogenetic signal, it is not immediately clear whether this is due to the trait itself having strong inertia, or if the trait is evolving in response to a phylogenetically structured variable. As we also obtain a measure of half-life for a response trait when testing it against a predictor, it is possible to examine which of these two phenomena underlies the phylogenetic signal in the trait. When a trait rapidly evolves in response to a phylogenetically structured variable (i.e. that which we test as a predictor), we may see a reduction in the half-life for the response trait as compared to the phylogenetic signal measured when we do not include the predictor in the regression (see Labra et al., 2009 for further explanations of this point).

SLOUCH analyses also consider standard errors of the input variables. For single sample species (i.e. for which standard errors could not be calculated), we took a conservative approach using the largest observed standard errors for other species. The largest standard error for (log) maternal body size in our data set is that of Lycia pomonaria (0.71), so we set the standard error for single sample species to 0.71. For (log) egg size, the largest standard error is that of Gymnoscelis rufifasciata (0.35), so we set the standard error for single sample species to 0.35.

SLOUCH returns an estimate of the similarity between the observed regression, or evolutionary regression (see Labra et al., 2009) between our response and predictor variables is to the so-called optimal regression. The optimal regression is the expected scenario in the hypothetical case when phylogenetic inertia in our response trait was absent (half-life = 0) (Hansen, 1997; Hansen et al., 2008; Labra et al., 2009). Comparing observed and optimal regressions indicates how close a response trait is to its optimal state and provides further indication as to how rapidly this response trait evolves.

Following the bivariate analysis of body and egg size, we modelled egg size as a dependent response variable against categorical ecological predictor variables (diet, habitat, overwintering stage). To test the influence of such ecological traits, the evolutionary histories of these ecological traits must be mapped on the phylogeny (Hansen, 1997; Butler & King, 2004). The evolutionary histories of all categorical variables were estimated by individually reconstructing them from the tip to the root of the tree using Fitch (Fitch, 1971) parsimony, allowing character state reversals along lineages. To evaluate the robustness of our results, both accelerated and delayed transformation optimization [ACCTRAN and DELTRAN (Swofford & Maddison, 1987)] of character states were tested. A third method, maximum likelihood ancestral state reconstructions (a more informed approach using phylogenetic branch length information) (Pagel, 1999; Davis et al., 2009), cannot be applied to mixed predictor models with continuous predictors because of current software limitations not allowing us to use ambiguous character states at nodes. In any case, the resulting maximum likelihood reconstructions we made range from 88% to 92% similar, depending on the variable, to those reconstructed using parsimony, and we do not expect that using maximum likelihood reconstructions would seriously alter the overall results of the analyses presented here.

All categorical variables were tested individually as predictors of egg size to see whether they could account for any of the variation observed in this trait. They were further tested as predictors against maternal body size as well to determine whether these variables are influencing maternal body sizes (i.e. is any phylogenetic signal seen in maternal body size explained by diet, habitat or overwintering state). In the present context, this was primarily important to evaluate the possibility that any covariation between maternal body size on egg size may be the result of a parallel adaptation of both these traits to particular selective environments.

We also wished to examine the effect of combinations of predictors, as the underlying influences in variation in egg size are widely acknowledged to be complex (e.g. Bernardo, 1996; Fox & Czesak, 2000). We might expect combinations of unrelated predictors to explain more of the variance in egg size than a single predictor alone (Labra et al., 2009). We therefore also ran models with combinations of body size, as a randomly changing, continuous predictor variable with individual fixed, categorical variables.

We compare how well different models explain the evolution of a response trait based on Akaike’s Information Criterion (AIC), as suggested by Butler & King (2004), supported by Freckleton (2009) and following Labra et al. (2009). Here we use AICc, which is a small sample adjustment of AIC (Burnham & Anderson, 1998). The variance explained for a given model is also provided by SLOUCH as a phylogenetically corrected R2 value.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

An important initial finding is that egg size exhibits very little phylogenetic signal (best model – 1.3% tree height), whereas we found evidence of considerable phylogenetic signal in maternal body size (best model – 21.9% tree height). Consequently, it is clearly more meaningful to model egg size as a response to optima influenced by maternal body size rather than vice versa (see materials and methods).

The best performing model explaining egg size evolution was found with maternal body size as a lone predictor (AICc = 80.121). The R2 of 56% indicates that maternal body size explains more than half of the variation in egg size optima. The best estimate of the evolutionary regression between egg and maternal body size shows a negative allometric relationship of egg size (log) on maternal body size (log) (slope 0.65 (SE 0.09), intercept -5.35 (SE 0.42), which implies a negative allometric relationship of egg size on body size. The optimal regression of log egg size on log maternal body size is almost identical to the observed regression, indicating that egg sizes are very close to their optimal states (Fig. 2). Other candidate models using single discrete predictors do not explain the variation in egg size so well and return much lower R2. Combining predictors (i.e. maternal body size plus one discrete variable) does not increase R2 substantially. For R2 values and AICc scores see Table 1.

image

Figure 2.  Optimal (dashed line) vs. observed evolutionary (dotted line) regressions of egg size on maternal body size. Optimal and evolutionary regressions are shown. The variance explained in egg size by maternal body size is R2 = 56.6%. We interpret that the absence of points in the top left corner indicates constraint imposed by maternal body size and the absence of points in the lower right corner should be a result of ecologically based selection against small eggs in larger moths.

Download figure to PowerPoint

Table 1.   Individual trait tests for phylogenetic signal in egg size and maternal body size (bold), and optimal regressions of different variables on egg size and maternal body size. A and D = ACCTRAN and DELTRAN character optimizations, respectively. INF = infinite half-life.
Trait (in bold) + predictorsHalf-life: % tree height (t1/2) [support region]R2AICc
  1. AIC, Akaike’s Information Criterion.

Egg size 1.3 [0.0–22.4]
+ Maternal body size0.4 [0.0–21.2]56.6080.121
+ Diet (poly/oligophagy)3.0 [0.0–21.2] (A)0.45 (A)118.055 (A)
1.1 [0.0–23.4] (D) 2.20 (D)116.496 (D)
+ Diet (tree/herb)4.0 [0.0–19.1] (A)9.12 (A)114.357 (A)
5.1 [0.0–19.1] (D)8.09 (D)112.588 (D)
+ Habitat5.1 [0.0–21.2] (A and D) 8.30 (A and D)112.462 (A and D)
+ Overwintering eggs1.1 [0.0–21.2] (A and D)0.01 (A and D)116.727 (A and D)
+ Maternal body size and diet (poly/oligophagy)0.4 [0.0–17.0] (D)56.50 (D)82.820 (D)
+ Maternal body size and diet (tree/herb)1.7 [0.0–17.0] (D)58.76 (D)80.555 (D)
+ Maternal body size and habitat0.4 [0.0–17.9] (A and D)57.20 (A and D) 81.940 (A and D)
+ Maternal body size and overwintering eggs0.4 [0.0–19.1] (A and D)56.66 (A and D)82.744 (A and D)
Maternal body size21.9 [0.0–85.1]
+ Diet (poly/oligophagy)3.2 [0.0–44.7] (A) 7.50 (A)135.730 (A)
22.3 [0.0–INF] (D) 0.15 (D)136.084 (D)
+ Diet (tree/herb)2.1 [0.0–46.8] (A) 6.80 (A)136.125 (A)
18.1 [0.0-INF] (D)2.24 (D)135.099 (D)
+ Habitat16.0 [0.0–57.4] (A and D) 6.60 (A and D)132.878 (A and D)
+ Overwintering eggs21.9 [0.0–INF] (A and D)3.16 (A and D)134.510 (A and D)

When maternal body size was treated as the response variable against ecological predictors, there are some large reductions in half-life with particular predictors (polyphagy vs. oligophagy ACCTRAN model – 3.2% tree height, tree vs. herb diet ACCTRAN model – 2.1% tree height) and some smaller reductions with others (Table 1). This could indicate that any phylogenetic signal we see in maternal body size is due to phylogenetic signal in these predictors. However, R2 values for best models indicate very little influence of these predictors on maternal body size. High AICc scores (i.e. low support) for these models support this claim (Table 1). Furthermore, where ACCTRAN models show obvious reductions in half-lives with diet-related predictors, DELTRAN models do not (Table 1). We believe that our results do not provide enough evidence against maternal body size being a phylogenetically inert trait. This scenario is also supported by some previous evidence, which shows body size in various taxa to exhibit substantial phylogenetic inertia (Diniz-Filho et al., 1998; Blomberg et al., 2003).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

We observed a positive correlation between maternal body size and egg size in geometrid moths. The relationship was found to be negatively allometric, which is in agreement with the pattern found in other phylogenetic comparative studies of Lepidoptera (García-Barros & Munguira, 1997; García-Barros, 2000), spiders (Marshall & Gittleman, 1994), birds (Olsen & Cockburn, 1993) and lizards (Bauwens & Diaz-Uriarte, 1997), perhaps demonstrating this as a widespread phenomenon. Quantitative comparisons across taxa are currently complicated by methodological differences among different studies, some of which do not consider phylogenetic effects (Rahn et al., 1975; Hines, 1982; Berrigan, 1991; Ryan & Lindeman, 2007). The increasing availability of phylogenies and respective computational methods allows one to expect rapid progress along these lines.

This study goes beyond just showing a positively correlated evolution of maternal body size and egg size. Having calculated phylogenetic half-lives of the two traits (much higher in body size than egg size), we can infer the likely direction of any evolutionary causality: egg size adapts to changes in body size and not vice versa. To our knowledge, for egg size and maternal body size, the causal direction of any potential adaptive relationship has never been explicitly tested before, and been just routinely assumed (Bernardo, 1996). Making this assumption, that is defining egg size as the response variable, enabled us to examine the relationship between these traits using the Ornstein-Uhlenbeck process-based comparative methods. The results indicate that evolutionary changes in maternal body size exert a large influence on the variation we see in egg size in this group (responsible for over half the variation). As we estimated a largely different magnitude of the phylogenetic signals for these two traits, we can also make the argument that the observed allometry is not related to a common underlying genetic mechanism, and therefore the hypothesis regarding genetic constraint appears unlikely. This is consistent with weak or no within-species phenotypic correlations between the two traits found in some other lepidopterans (Wiklund et al., 1987; Tammaru et al., 2002; Bauerfeind & Fischer, 2008).

Furthermore, we find no evidence that the evolution of egg size or maternal body size is strongly related to ecological factors previously hypothesized to affect optima for these traits: there was no effect of larval host range, tree- vs. herb feeding in larvae, habitat or overwintering stage. Therefore, we tend to exclude the possibility that the relationship we see between maternal body size and egg size is indirect (i.e. mediated by an independent adaptation of both of these traits to a third factor). Not only does this rule out that these factors strongly influence egg size variation, but it also gives no support for respective hypotheses suggesting that they influence body size in Lepidoptera (e.g. Nakasuji, 1987; Reavey, 1992; Braby, 1994). Combining maternal body size as a predictor of egg size with independent ecological predictors does not explain much extra variation in egg size than does maternal body size alone.

As maternal body size explains just 56% of the variation in egg size, selective factors other than those considered in this study should play a role. A promising path may be to relate egg size to egg production strategies (capital vs. income breeding, pro- vs. synovigeny: Jervis & Ferns, 2004; Javoišet al., 2011), but such analyses are currently hindered by scarcity of comparable data available for a sufficient number of species. Whether multiple ecological factors in combination play a larger role in determining egg size variation cannot be tested due to current software limitations and testing this must wait until the relevant software becomes available.

Without any evidence for another factor playing a major role in egg size variation, we must ask how maternal body size determines egg size. Our finding that the phylogenetic half-lives of these traits are markedly different suggests that egg size changes rapidly follow maternal body size changes. The closeness of the actual regression of the values of egg size on maternal body size to the optimal regression (Fig. 2) further demonstrates that egg size is a highly responsive trait. Although the comparative methods based on an Ornstein-Uhlenbeck model of evolution were designed to model adaptive evolution (Hansen, 1997; Butler & King, 2004; Hansen et al., 2008), hypotheses of constraint (in this context: evolutionary forces which are not ecological in their nature) should always be considered as an alternative explanation to those of adaptation (Maynard Smith et al., 1985). Rather than egg size changes occurring primarily as an adaptation to the external environment, our results suggest that maternal body size acts as an anatomical constraint that affects the adaptive optima for egg size. In other words, it may be complicated for small moths to lay large eggs for reasons that are not ecological in their nature. When maternal body size increases (and therefore the constraint is lessened), egg size is further able to increase and rapidly follows suit.

Indeed, it is rather intuitive to see that some anatomical and physiological constraints on large eggs in small insects are inevitable (e.g. caused by limits on size of some part of the reproductive organs, etc.). Such constraints would explain why, in our study, the top left corner of the graph in Fig. 2 has no data points. However, it is much less clear why some large mothers do not lay a high number of small eggs (i.e. why is the lower right corner of Fig. 2 devoid of data points?). If egg size was driven by external ecological selective factors to a considerable extent, and given the ecological diversity of geometrid moths, it should be reasonable to assume that the optimal solution to the offspring quality (size)-quantity trade-off should sometimes substantially differ, even for similarly sized moths. As this appears not to be the case, we are inclined to infer that there is some rather universal ecologically based selection towards larger egg size in geometrid moths: it is advantageous to lay as large eggs as permitted by maternal body size. The exact nature of the involved selective forces remains to be elucidated.

On a broad scale, this study indicates the potential of the most recently developed OU-based phylogenetic comparative methods to enhance our understanding of relationships between life history traits. With the novel and rapidly developing OU-based methods, we can begin to understand the nature of the phylogenetic signal observed in traits of interest – i.e. which of those respond quickly (be that adaptation to the environment or following some constraining physiological factor) or slowly (e.g. due to phylogenetic inertia) – and consider the extent of adaptive or constraint-based relationships between life history traits. We eagerly await extensions to currently available software, which will allow more in-depth exploration of the role of categorical variables, which in the current context will allow us to assess in greater depth the role of ecological habit on life history evolution.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

The study was supported by Estonian Science Foundation Grant 7699, targeted financing project SF0180122s08 and by the European Union through the European Regional Development Fund (Center of Excellence FIBIR). We thank Freerk Molleman and Ants Kaasik for useful comments and advice.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information
  • Avezedo, R.B.R., French, V. & Partridge, L. 1997. Life-history consequences of egg size in Drosophila Melanogaster. Am. Nat. 150: 250282.
  • Barbosa, P. & Capinera, J.L. 1978. Population quality, dispersal and numerical change in the gypsy moth, Lymantria dispar (L.). Oecologia 36: 203209.
  • Barbosa, P., Krischik, V. & Lance, D. 1989. Life-history traits of forest-inhabiting flightless Lepidoptera. Am. Midl. Nat. 122: 262274.
  • Bauerfeind, S.S. & Fischer, K. 2008. Maternal body size as a morphological constraint on egg size and fecundity in butterflies. Basic Appl. Ecol. 9: 443451.
  • Bauwens, D. & Diaz-Uriarte, R. 1997. Covariation of life-history traits in lacertid lizards: a comparative study. Am. Nat. 149: 91111.
  • Bernardo, J. 1996. The particular maternal effect of propagule size, especially egg size: patterns, models, quality of evidence and interpretations. Am. Zool. 36: 216236.
  • Berrigan, D. 1991. The allometry of egg size and number in insects. Oikos 60: 313321.
  • Blackburn, T.M. 1991. Evidence for a fast slow continuum of life-history traits among parasitoid Hymenoptera. Funct. Ecol. 5: 6574.
  • Blomberg, S.P., Garland, T. & Ives, A.R. 2003. Testing for phylogenetic signal in comparative data: behavioral traits are more labile. Evolution 57: 717745.
  • Braby, M.F. 1994. The significance of egg size variation in butterflies in relation to hostplant quality. Oikos 71: 119129.
  • Brehm, G. & Fiedler, K. 2004. Bergmann’s rule does not apply to geometrid moths along an elevational gradient in an Andean montane rain forest. Glob. Ecol. Biogeogr. 13: 714.
  • Burnham, K.P. & Anderson, D.R. 1998. Model Selection and Inference: A Practical Information-Theoretic Approach. Springer, New York.
  • Butler, M.A. & King, A.A. 2004. Phylogenetic comparative analysis: a modeling approach for adaptive evolution. Am. Nat. 164: 683695.
  • Buxton, P.A. 1955. The Natural History of Tsetse Flies. Lewis & Co., London.
  • Byrne, P.G., Simmons, L.W. & Roberts, J.D. 2003. Sperm competition and the evolution of gamete morphology in frogs. Proc. R. Soc. Lond. B Biol. Sci. 270: 20792086.
  • Carriere, Y. & Roff, D.A. 1995. The evolution of offspring size and number: a test of the Smith-Fretwell model in three species of crickets. Oecologia 102: 389396.
  • Cooke, S.J., Philipp, D.P., Wahl, D.H. & Weatherhead, P.J. 2006. Energetics of parental care in six syntopic centrarchid fishes. Oecologia 148: 235249.
  • Crespi, B.J. & Teo, R. 2002. Comparative phylogenetic analysis of the evolution of semelparity and life history in salmonid fishes. Evolution 56: 10081020.
  • Czesak, M.E. & Fox, C.W. 2003. Evolutionary ecology of egg size and number in a seed beetle: genetic trade-off differs between environments. Evolution 57: 11211132.
  • Davis, R.B., Baldauf, S.L. & Mayhew, P.J. 2009. Eusociality and the success of the termites: insights from a supertree of dictyopteran families. J. Evol. Biol. 22: 17501761.
  • Diniz-Filho, J.A.F., de Sant’ana, C.E.R. & Bini, L.M. 1998. An Eigenvector Method for Estimating Phylogenetic Inertia. Evolution 52: 12471262.
  • Dixon, A.F.G. & Hemptinne, J.-L. 2001. Body size distribution in predatory ladybird beetles reflects that of their prey. Ecology 82: 18471856.
  • Drummond, A.J. & Rambaut, A. 2007. BEAST: bayesian evolutionary analysis by sampling trees. BMC Evol. Biol. 7: 214.
  • Dyke, G.J. & Kaiser, G.W. 2010. Cracking a developmental constraint: egg size and bird evolution. Rec. Aust. Mus. 62: 207216.
  • Felsenstein, J. 1985. Phylogenies and the comparative method. Am. Nat. 125: 115.
  • Ficetola, G.F. & de Bernardi, F. 2009. Offspring size and survival in the frog Rana latasei: from among-population to within-clutch variation. Biol. J. Linn. Soc. 97: 845853.
  • Fischer, K., Zwaan, B.J. & Brakefield, P.M. 2002. How does egg size relate to body size in butterflies? Oecologia 131: 375379.
  • Fischer, K., Brakefield, P.M. & Zwaan, B.J. 2003. Plasticity in butterfly egg size: why large offspring at lower temperatures? Ecology 84: 31383147.
  • Fitch, W.M. 1971. Toward defining the course of evolution: minimum change for a specific tree topology. Syst. Zool. 20: 406416.
  • Forbes, E.L., Preston, C.D. & Lokman, P.M. 2010. Zebrafish (Danio rerio) and the egg size versus egg number trade off: effects of ration size on fecundity are not mediated by orthologues of the Fec gene. Reprod. Fertil. Dev. 22: 10151021.
  • Fox, C.W. 1994. The influence of egg size on offspring performance in the seed beetle Callosobruchus maculatus. Oikos 71: 321325.
  • Fox, C.W. & Czesak, M.E. 2000. Evolutionary ecology of progeny size in arthropods. Annu. Rev. Entomol. 45: 341369.
  • Freckleton, R.P. 2009. The seven deadly sins of comparative analysis. J. Evol. Biol. 22: 13671375.
  • García-Barros, E. 1992. Evidence for geographic variation of egg size and fecundity in a satyrine butterfly, Hipparchia semele (Lepdioptera, Nymphalidae-Satyrinae). Graellsia 48: 4552.
  • García-Barros, E. 1994. Egg size variation in European Satyrinae butterflies (Nymphalidae, Satyrinae). Biol. J. Linn. Soc. 51: 309324.
  • García-Barros, E. 2000. Body size, egg size, and their interspecific relationships with ecological and life history traits in butterflies (Lepidoptera: Papilionoidea, Hesperioidea). Biol. J. Linn. Soc. 70: 251284.
  • García-Barros, E. 2002. Taxonomic patterns in the egg size to body size allometry of butterflies and skippers (Papilionoidea and Hesperiidae). Nota Lepidopterol. 25: 161175.
  • García-Barros, E. & Munguira, M.L. 1997. Uncertain branch lengths, taxonomic sampling error, and the egg to body size allometry in temperate butterflies (Lepidoptera). Biol. J. Linn. Soc. 61: 201221.
  • Hansen, T.F. 1997. Stabilizing selection and the comparative analysis of adaptation. Evolution 51: 13411351.
  • Hansen, T.F. & Orzack, S.H. 2005. Assessing current adaptation and phylogenetic inertia as explanations of trait evolution: the need for controlled comparisons. Evolution 59: 20632072.
  • Hansen, T.F., Pienaar, J. & Orzack, S.H. 2008. A comparative method for studying adaptation to a randomly evolving environment. Evolution 62: 19651977.
  • Hausmann, A. 2001. The Geometrid Moths of Europe. Volume 1. Introduction Archiearinae, Orthostixinae, Desmobathrinae, Alsophilinae, Geometrinae. Apollo Books, Stenstrup, Denmark.
  • Hausmann, A. 2004. The Geometrid Moths of Europe Volume 2: Sterrhinae. Apollo Books, Stenstrup, Denmark.
  • Hines, A.H. 1982. Comparative growth-patterns in the life histories of crabs. Am. Zool. 22: 938.
  • Hinton, H.E. 1981. Biology of Insect Eggs, Vol. 1. Pergamon Press, Oxford.
  • Honěk, A. 1992. Female size, reproduction and progeny size in Pyrrhocoris apterus (Heteroptera, Pyrrhocoridae). Acta Entomol. Bohemoslov. 89: 169178.
  • Javois̆, J., Molleman, F. & Tammaru, T. 2011. Quantifying income breeding: using geometrd moths as an example. Entomol. Exp. Appl. 139: 187196.
  • Jervis, M.A. & Ferns, P.N. 2004. The timing of egg maturation in insects: ovigeny index and initial egg load as measures of fitness and of resource allocation. Oikos 107: 449461.
  • Johnson, N.F., Rawlins, J.E. & Pavuk, D.M. 1987. Host-related antennal variation in the polyphagous egg parasite Telenomus alsophilae (Hymenoptera: Scelionidae). Syst. Entomol. 12: 437447.
  • Kinnison, M.T., Unwin, M.J., Hendry, A.P. & Quinn, T.P. 2001. Migratory costs and the evolution of egg size and number in introduced and indigenous salmon populations. Evolution 55: 16561667.
  • Koch, M. 1984. Wir Bestimmen Schmetterlinge. Neumann-Neudamm, Berlin, Germany.
  • Labra, A., Pienaar, J. & Hansen, T.F. 2009. Evolution of thermal physiology in Liolaemus lizards: adaptation, phylogenetic inertia, and niche tracking. Am. Nat. 174: 204220.
  • Larson, K. & Forslund, P. 1992. Genetic and social inheritance of body and egg size in the barnacle goose (Branta leucopsis). Evolution 46: 235244.
  • Leraut, P. 2009. Moths of Europe Volume 2. Geometrid Moths. NAP Editions, Verrières le Buisson, France.
  • Marshall, N.B. 1953. Egg size in arctic, antarctic and deep sea fishes. Evolution 7: 328341.
  • Marshall, S.D. & Gittleman, J.L. 1994. Clutch size in spiders – is more better? Funct. Ecol. 8: 118124.
  • Maynard Smith, J., Burian, R., Kauffman, S., Alberch, P., Campbell, J., Goodwin, B. et al. 1985. Developmental constraints and evolution. Q. Rev. Biol. 60: 265287.
  • Minet, J. & Scoble, M.J. 1999. The Drepanoid/Geometroid assemblage. In: Handbook of Zoology, Vol. IV (35), Lepidoptera, Moths and Butterflies, 1: Evolution, Systematics and Biogeography (N.P. Kristensen, ed.), pp. 301320. Walter de Gruyter, Berlin, Germany.
  • Moran, A.L. & Emlet, R.B. 2001. Offspring size and performance in variable environments: field studies on a marine snail. Ecology 82: 15971612.
  • Nakasuji, F. 1987. Egg size of skippers (Lepidoptera: Hesperiidae) in relation to their host specificity and to leaf toughness of host plants. Ecol. Res. 2: 175183.
  • Oberhauser, K.S. 1997. Fecundity, lifespan and egg mass in butterflies: effects of male-derived nutrients and female size. Funct.Ecol. 11: 166175.
  • Olsen, P.D. & Cockburn, A. 1993. Do large females lay small eggs? Sexual dimorphism and the allometry of egg and clutch volume Oikos 66: 447453.
  • Õunap, E., Viidalepp, J. & Saarma, U. 2008. Systematic position of Lythriini revised: transferred from Larentiinae to Sterrhinae (Lepidoptera, Geometridae). Zool. Scr. 37: 405413.
  • Õunap, E., Javoiš, J., Viidalepp, J. & Tammaru, T. 2011. Phylogenetic relationships of selected European Ennominae (Lepidoptera, Geometridae). Eur. J. Entomol. 108: 267273.
  • Pagel, M. 1999. Inferring the historical patterns of biological evolution. Nature 401: 877884.
  • Palmer, J.O. 1985. Life-history consequences of body-size variation in the milkweed leaf beetle Labidomera clivicollis (Coleoptera: Chrysomelidae). Ann. Entomol. Soc. Am. 78: 603608.
  • Poulin, R. & Hamilton, W.J. 1997. Ecological correlates of body size and egg size in parasitic Ascothoracida and Rhizocephala (Crustacea). Acta Oecol. 18: 621635.
  • R Development Core Team (2005). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org
  • Rahn, H., Paganelli, C.V. & Ar, A. 1975. Relation of avian egg weight to body weight. Auk 92: 750765.
  • Reavey, D. 1992. Egg size, 1st instar behavior and the ecology of Lepidoptera. J. Zool. 227: 277297.
  • Ryan, K.M. & Lindeman, P.V. 2007. Reproductive allometry in the common map turtle, Graptemys geographica. Am. Midl. Nat. 158: 4959.
  • Salthe, S.N. 1969. Reproductive modes and the number and sizes of ova in the Urodeles. Am. Midl. Nat. 81: 467490.
  • Schall, J.J. 1978. Reproductive strategies in sympatric whiptail lizards (Cnemidophorus): two parthenogenetic and three bisexual species. Copeia 1: 108116.
  • Smith, C.C. & Fretwell, S.D. 1974. The optimal balance between size and number of offspring. Am. Nat. 108: 499506.
  • Snäll, N., Tammaru, T., Wahlberg, N., Viidalepp, J., Ruohomäki, K., Savontaus, M.-L. et al. 2007. Phylogenetic relationships of the tribe Operophterini (Lepidoptera, Geometridae): a case study of the evolution of female flightlessness. Biol. J. Linn. Soc. 92: 241252.
  • Swofford, D.L. & Maddison, W.P. 1987. Reconstructing ancestral character states under Wagner parsimony. Math. Biosci. 87: 199229.
  • Tammaru, T., Kaitaniemi, P. & Ruohomäki, K. 1995. Oviposition choices of Epirrita autumnata (Lepidoptera: Geometridae) in relation to its eruptive population dynamics. Oikos 74: 296304.
  • Tammaru, T., Esperk, T. & Castellanos, I. 2002. No evidence for costs of being large in females of Orgyia spp. (Lepidoptera, Lymantriidae): larger is always better. Oecologia 133: 430438.
  • Van Buskirk, J. & Crowder, L.B. 1994. Life-history variation in marine turtles. Copeia 1: 6681.
  • Viidalepp, J., Tammaru, T., Snäll, N., Ruohomäki, K. & Wahlberg, N. 2007. Cleorodes, Warren 1894 does not belong to the tribe Boarmiini (Lepidoptera: Geometridae). Eur. J. Entomol. 104: 303309.
  • Wahlberg, N., Snäll, N., Viidalepp, J., Ruohomäki, K. & Tammaru, T. 2010. The evolution of female flightlessness among Ennominae of the Holarctic forest zone (Lepidoptera, Geometridae). Mol. Phylogenet. Evol. 55: 929938.
  • Walz, N., Sarma, S.S.S. & Benker, U. 1995. Egg size in relation to body size in rotifers: an indication of reproductive strategy? Hydrobiologia 313: 165170.
  • Wasserman, S.S. & Mitter, C. 1978. The relationship of body size to breadth of diet in some Lepidoptera. Ecol. Entomol. 3: 155160.
  • Wellington, G. 1965. Some maternal influences on progeny quality in the western tent caterpillar Malacosoma pluviale (Dyar). Can. Entomol. 97: 114.
  • Wiklund, C. & Karlsson, B. 1984. Egg size variation in satyrid butterflies: adaptive vs historical, “Bauplan”, and mechanistic explanations. Oikos 43: 391400.
  • Wiklund, C., Karlsson, B. & Forsberg, J. 1987. Adaptive versus constraint explanations for egg-to-body size relationships in two butterfly families. Am. Nat. 130: 828838.
  • Williams, T.D. 1994. Intraspecific variation in egg size and egg composition in birds: effects on offspring fitness. Biol. Rev. 69: 3559.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  9. Supporting Information

Appendix S1 Phylogenetic analysis

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
JEB_2420_sm_supplementarymaterial.doc27KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.