Factors promoting maternal trophic egg provisioning in non-eusocial animals

Abstract

The adaptive function of trophic egg-laying is generally regarded as extended parental investment to the offspring. However, the evolutionary factors promoting trophic egg-laying are still unclear, because the amount of maternal investment per offspring should be ideally equal between smaller offspring with trophic eggs and larger offspring without any additional investment. Several authors have suggested that trophic egg-laying should evolve only when egg size is constrained, but this hypothesis has not been evaluated. We investigated the evolutionary mechanisms of trophic egg-laying by two different approaches. First, we evaluated morphological constraints on egg size in two sibling ladybird species, Harmonia axyridis, which is known to produce trophic eggs, and H. yedoensis. Second, we theoretically predicted the optimal proportion of trophic eggs to total eggs and egg size in relation to environmental heterogeneity, predictability of environmental quality, and investment efficiency of trophic egg consumption. The intra- and interspecific morphological comparisons suggest that morphological constraints on the evolutionary determination of egg size are weak at best in the two ladybird species. Moreover, we theoretically showed that small egg size and trophic egg-laying are favoured in heterogeneous environments when mothers cannot adjust egg size plastically. We also showed that even a small reduction in investment efficiency makes a trophic egg strategy unlikely, despite relatively high environmental predictability. We conclude that trophic egg provisioning may be a flexible maternal adaptation to a highly heterogeneous environment rather than a response to a morphological constraint on egg size.

Introduction

Trophic eggs (also called nurse eggs) are non-developing eggs or egg-like structures produced for offspring consumption (Crespi 1992). Strictly speaking, trophic egg-laying is an evolved maternal phenotype, not simply the unavoidable production of non-developing eggs that happen to be eaten by offspring (Crespi 1992; Perry and Roitberg 2006). Trophic egg-laying occurs in diverse animal taxa with various parental care systems (e.g., non-social and eusocial arthropods, gastropods, amphibians, fishes; reviewed by Perry and Roitberg 2006), although it is often confined to only a few species within a taxonomic group (Elgar and Crespi 1992). In general, the adaptive function of trophic egg-laying is regarded as extended parental investment to the offspring (Alexander 1974; Polis 1984). However, the amount of maternal investment per offspring should be ideally equal between smaller offspring with trophic eggs and larger offspring without any additional parental investment (Baur 1990; Dixon 2000). Therefore, evolutionary factors promoting trophic egg-laying, instead of larger offspring size, are still not understood.

Several studies have suggested that trophic egg-laying should evolve only when egg size is constrained (Alexander 1974; Polis 1984; Mock and Parker 1997; Dixon 2000). Morphological constraints on egg size such as the size of the ovipositor or pelvic aperture may prevent small females from producing large eggs (Congdon and Gibbons 1987) and thus may lead to the evolution of trophic egg-laying. Moreover, other hypotheses based on the density effect of competing offspring (Parker and Begon 1986) and the physiological constraints (Sakai and Harada 2001) also explain why larger mothers produce larger offspring. In fact, many field studies of diverse animal taxa have documented positive correlations between maternal size and offspring size (Fox and Czesak 2000). However, no studies have evaluated whether mothers that produce trophic eggs exhibit such constraints on maternal body size.

Adaptive mechanisms that might promote trophic egg-laying should be considered in situations where morphological constraints on egg size are not critical. Trophic egg-laying occurs in several groups of eusocial Hymenoptera (Sakagami 1982; Hölldobler and Wilson 1990; Crespi 1992), but it is difficult to examine the adaptive significance of trophic eggs in eusocial systems because complicated conflicts among colony members may obscure the origin and evolution of trophic eggs (Crespi 1992). By contrast, in non-eusocial taxa, the adaptive function of trophic eggs for offspring survival has been examined by focusing on environmental heterogeneity (Kudo and Nakahira 2005, Perry and Roitberg 2005a). In fact, studies have documented that females of some non-eusocial species that face highly heterogeneous environments adopt a trophic egg strategy (e.g., Crump 1981; Kudo and Nakahira 2004). However, maternal fitness between the two strategies, (1) producing small offspring with trophic eggs and (2) producing large offspring without any additional investment, has not been compared explicitly. For such a comparison, a model that can predict which strategy is evolutionarily stable in a heterogeneous environment needs to be developed.

Such a model must take into account the unavoidable costs and limitations that are likely to accompany a trophic egg strategy in a heterogeneous environment. As in any strategy involving adaptive phenotypic plasticity (Berrigan and Scheiner 2004; Marshall and Uller 2007), cues that reliably predict future environment conditions must be present for flexible trophic egg provisioning to evolve. However, the environment that the offspring will face is not always predictable, especially in species in which offspring and adult individuals occupy different niches (Werner and Gilliam 1984; Fischer et al. 2011). Although to reduce the level of uncertainty, mothers can collect information that will be useful in making provisioning decisions (Dall et al. 2005), in a variable environment, a certain level of uncertainty is likely to persist. Moreover, even when mothers can collect information accurately, it may be difficult to compensate for environmental quality in through provisioning without specialized morphology and physiology for trophic egg production, such as distinct ovariole structure and cellular development. Indeed, in the predatory ladybird Harmonia axyridis which lacks specialized trophic egg structure (Osawa and Yoshinaga 2009), mothers can manipulate the proportion of trophic eggs depending on the prey availability, but the ratio of trophic to viable eggs is variable even in the highly standardized laboratory conditions (Perry and Roitberg 2005a). As a result, the possibility exists that mothers will provision the ‘wrong’ amount of resources to their offspring. Furthermore, consumption of trophic eggs is likely to involve some waste of maternal resources (Elgar and Crespi 1992). However, no study incorporating such costs and limitations has thus far examined the conditions that favour the trophic egg-laying.

In this study, we investigated the evolutionary mechanisms that favour trophic egg-laying by making morphological comparisons and by mathematical modelling. First, we compared egg size and maternal body size both intra- and interspecifically in two sibling ladybird species, Harmonia axyridis Pallas and Harmonia yedoensis Takizawa. Adult body size is quite similar in these two ladybird species (Sasaji 1998), and the hatched larvae of both species eat clusters of sibling eggs, both undeveloped eggs and developing eggs with delayed hatching (Kawai 1978; Osawa 1992a; Osawa and Ohashi 2008). The sibling egg consumption by hatchlings can be regarded as an adaptive maternal phenotype in both H. axyridis (Perry and Roitberg 2005a) and H. yedoensis (Noriyuki et al. 2011). If morphological constraints on egg size exist, a positive correlation between egg size and maternal body size would be expected within each species, because smaller mothers cannot produce larger eggs (Fox and Czesak 2000; Fischer et al. 2002; Noriyuki et al. 2010). In addition, if morphological constraints prevent H. axyridis females from producing large eggs, then the similar-sized H. yedoensis females should not be able to produce eggs larger than those of H. axyridis. Furthermore, in females of both species we also examined the ovariole number, which is an important determinant of egg size for a given maternal body size in insects (Gilbert 1990; Stewart et al. 1991a). Thus, we evaluated the role of morphological constraints by comparing and assessing morphological traits in these two sibling ladybird species.

Second, we constructed a mathematical model to predict the optimal proportion of trophic eggs and egg size that mothers should produce. Under spatially and temporally heterogeneous environments for offspring survival, we investigated how reliable the environmental cues available to the mother have to be for a trophic egg strategy to be favoured by selection. Additionally, we incorporated the fact that a certain proportion of maternal resources provided as trophic eggs are not consumed by the offspring and examined whether natural selection favours trophic egg-laying despite its cost. Finally, on the basis of the results of these different two approaches and the findings of previous studies, we propose a reasonable explanation for the widespread occurrence of trophic egg-laying in non-eusocial animals.

Methods

Morphological measurements

The generalist predator H. axyridis and the specialist predator H. yedoensis are sibling species with sympatric distributions in central Japan (Noriyuki et al. 2011). Females of the two species produce undeveloped eggs which are consumed by the sibling hatchlings (Osawa and Ohashi 2008). However, the precise developmental mechanisms regulating the production of undeveloped eggs are not fully understood. Indeed, it is possible that gamete incompatibility and sperm limitation cause the failure of fertilization (e.g., Wedell et al. 2002). Moreover, eggs can be infected by male-killing bacteria and killed male embryos appear as infertile eggs in both H. axyridis (Majerus et al. 1998) and H. yedoensis (N. Osawa, unpublished data), although uninfected females also produce certain proportion of undeveloped eggs. The absence of micropyles is one criterion used to categorize an egg as trophic in the Hymenotera and Heteroptera (e.g., Gobin et al. 1998; Kudo et al. 2006), but in H. axyridis, micropyles are present in the shells of both developing and undeveloped eggs (Osawa and Yoshinaga 2009). In addition, there is no special feature of the spatial distribution of undeveloped eggs within the clutch in H. axyridis (Perry and Roitberg 2005a). However, the proportion of undeveloped eggs increases when H. axyridis mothers face the low resource availability (Perry and Roitberg 2005a) and consumption of undeveloped eggs enhance the survival of hatchlings especially when aphid density is low (Osawa 1992a), in a manner consistent with a definition of adaptive trophic egg provisioning (Perry and Roitberg 2006). Although H. yedoensis mothers have not been definitely proved to produce trophic eggs in the strict sense (cf. Perry and Roitberg 2006), the consumption of undeveloped eggs greatly enhances prey capture performance in H. yedoensis hatchlings, suggesting that the production of undeveloped eggs has evolved as an adaptive maternal phenotype (Noriyuki et al. 2011). Therefore, even though precise proximate mechanisms have not been identified, the production of undeveloped eggs in H. axyridis and H. yedoensis can be regarded as an adaptive maternal strategy for the offspring survival.

We collected 10 H. axyridis adults at the Botanical Garden of Kyoto University, Kyoto (135°47′E, 35°02′N), and 25 egg clusters of H. yedoensis at Hieidaira, Shiga (135°83′E, 35°02′N), in May 2008. We obtained eggs from the adults and the egg clusters, and reared the resulting offspring at each stage in plastic cups (13 cm wide, 10 cm high) to the adult stage at 25 °C, 16L:8D, and about 70 % relative humidity. We provided the larvae with a surplus of frozen Ephestia kuehniella Zeller eggs (Beneficial Insectary, Inc., Redding, CA, USA). We randomly chose 54 newly emerged and unmated individuals (27 females and 27 males) of H. axyridis and 48 newly emerged and unmated individuals (24 females and 24 males) of H. yedoensis from the stock for the experiment. We used first-generation offspring because (1) there were not enough adults of H. yedoensis in the original field collection for a valid statistical analysis, and (2) it allowed us to use larval morphology to confirm the identities of the two species, which are almost impossible to distinguish on the basis of adult morphology (Sasaji 1998).

To obtain eggs from the first-generation adults, we reared mated females individually with a surplus of frozen eggs at 25 °C, 16L:8D, and about 70 % relative humidity. We used 10 eggs from each of five different clutches laid by each female for data analysis, except if the number of eggs in a clutch was <10, then all eggs in that clutch were measured. In the case of a female that laid fewer than five clutches, we measured 10 eggs from each clutch that she laid. We measured both egg height (h) and width (r) under a stereo microscope (Carl Zeiss® SV-11 APO) to the nearest 0.025 mm. We estimated egg size as the egg volume calculated using the formula hr 2 π/6 (mm3; Takakura 2004). We measured body length with a slide calipers to the nearest 0.01 mm and used as maternal body size. After a female died or had laid five clutches, we examined her number of ovarioles under the stereo microscope.

Statistical analysis

We took into account the fact that the sizes of eggs from the same clutch or laid by the same mothers are not statistically independent. To test for a correlation between maternal body size and egg size, we adopted a regression model with more than one value of the dependent variable (egg size) per value of the independent variable (maternal body size; Sokal and Rohlf 1995). We used nested analysis of variance (nested ANOVA) with mothers within ladybird species, and with clutches within mothers, to compare egg size between ladybird species. We analyzed the effects of maternal body size and ladybird species on ovariole number by analysis of covariance (ANCOVA). All statistical analyses were carried out with JMP® 7 (SAS Institute Japan).

The model

We present the simplest theoretical framework for the evolution of trophic egg-laying that incorporates environmental heterogeneity, environmental predictability, and investment efficiency of trophic eggs. We assume that mothers cannot adjust egg size plastically. This assumption is applicable to many kinds of animals because ovariole or pelvic aperture size should remain unchanged in a given individual female. In fact, egg size appears to be inflexible within individual females in many animals, such as land snails (Baur 1988; Baur and Raboud 1988), insects (Stewart et al. 1991a, b; Dixon and Guo 1993; Soares et al. 2001), and birds (Christians 2002), and this inflexibility may be tied to ovariole or pelvic aperture size. Even though some species are able to manipulate egg size in response to the quality of the oviposition site (Leather and Burnand 1987; Fox et al. 1997; Mizumoto and Nakasuji 2007), plastic adjustment of egg size is uncommon in animal taxa and the degree of the egg size variation seems to be highly constrained (e.g., Kawecki 1995). In addition to morphological factors, physiological mechanisms of oogenesis may also constrain the immediate adjustment of egg size. Therefore, we assume instead that egg size can evolve to an optimal value and that mothers can produce trophic eggs to deal with an adverse environment. We assume that trophic eggs and viable eggs are the same size, because no general size difference trend has been reported. Moreover, we assume that trophic and viable eggs provide equal food quality for offspring survival, although possible differences in chemical composition and function between these eggs have not been examined. For simplicity, we assume that the total amount of maternal resources that can be allocated to the offspring as viable eggs and trophic eggs is the same for all females and fixed to 1.

To model the simplest possible variable environment, we examine an environment with only two states: good and poor. The relationship between investment per offspring and offspring fitness differs between the two environments because of biotic or abiotic factors. Moreover, we do not distinguish between spatial and temporal variation. We specify that the minimum viable offspring size is smaller in the good environment than in the poor environment, because an offspring should require fewer resources to become established in the more favourable environment (Fox et al. 1997). We also assume that, for offspring of a given size, offspring fitness is greater in the good environment than in the poor environment (McGinley et al. 1987). Furthermore, we assume that low levels of maternal investment result in zero fitness for offspring because they need a threshold amount of resource to survive, and that fitness approaches an asymptote at high levels of parental investment because offspring cannot make full use of excess resources (Smith and Fretwell 1974; Parker et al. 2002). Thus, we describe the relationship between offspring fitness S and viable egg size in the good environment by S G(e) = 1 – (1/e)2 and in the poor environment by S P(e) = 1 – (k/e)2, where the subscripts G and P indicate the good and the poor environment, respectively, e is viable egg size and trophic egg size, and k is a constant that specifies the difference in quality between the two environments. To make the labels ‘good’ and ‘poor’ biologically feasible, we assume k > 1 such that offspring survive better in a good environment. We assume that individual females experience the two habitat types, good and poor, in the proportion p and 1 − p, respectively.

We define environmental predictability, q, as the probability that maternal assessment of the environmental quality is correct. Specifically, we assume that when mothers incorrectly assess a poor environment as a good environment, then they do not provision trophic eggs although the offspring may need them to survive. Conversely, when maternal assessment of the good environment is wrong, then mothers may provision unneeded trophic eggs, causing per offspring maternal investment to exceed the optimal value (Table 1). For simplicity, we assume that predictability q is constant across environmental situations. If q = 1, then mothers can assess the environmental quality perfectly and provision the optimal proportion of trophic eggs in each environment; in contrast, if q = 0.5, then mothers provision trophic eggs with a probability of 0.5 irrespective of the environmental quality. Because q < 0.5 is not realistic, we consider only situations with 0.5 ≤ q ≤ 1 in the analysis.

Table 1 Relationship between environmental conditions and maternal investment

From the above assumptions it follows that the ratio of trophic eggs to the total amount of maternal investment is given by p(1 − q)t + (1 − p)qt = t(p + q – 2pq), where t is the ratio of trophic eggs to the amount of maternal investment when maternal assessment is wrong in the good environment or correct in the poor environment (that is, the amount of resource when mothers ‘evaluate’ the environment as poor; Table 1). Similarly, the number of viable eggs in an environment that mothers evaluate as poor is given by {p(1 − q)(1 − t) + (1 − p)q(1 − t)}/e = {(1 − t)(p + q − 2pq)}/e. Hence, per offspring maternal investment in an environment that mothers evaluate as poor is given by e + t(p + q − 2pq)δ ÷ {(1 − t)(p + q − 2pq)}/e = {1 + δt/(1 − t)}e, where δ is the investment efficiency of trophic egg consumption, defined as the proportion of the trophic egg amount consumed by the offspring. Because the total amount of maternal investment is fixed to 1, the amount of maternal investment that is provided as viable eggs in both environments is given by 1 − t(p + q − 2pq). Therefore, maternal fitness can be described as the product of offspring number times the probability of offspring survival in each environment:

$$ W(e,t) = \frac{1 - t(p + q - 2pq)}{e}\left\{ {p\left( {qS_{\text{G}} [e] + (1 - q)S_{\text{G}} \left[ {\left( {1 + \frac{\delta t}{1 - t}} \right)e} \right]} \right) + (1 - p)\left( {qS_{\text{P}} \left[ {\left( {1 + \frac{\delta t}{1 - t}} \right)e} \right] + (1 - q)S_{\text{P}} [e]} \right)} \right\}. $$

Results

Morphological comparisons

Mean female body length was not significantly different between H. axyridis (mean ± SE 7.03 ± 0.09 mm, n = 27) and H. yedoensis (7.24 ± 0.10 mm, n = 24; Student’s t test, t 49 = −1.67, P = 0.10). Female body length and egg volume were not significantly related in either species (linear regression analysis, H. axyridis: F 1,25 = 0.16, P = 0.69, R 2 = 0.003; H. yedoensis: F 1,22 = 0.25, P = 0.62, R 2 = 0.005; Fig. 1). Mean egg volume was significantly smaller in H. axyridis (mean ± SE 0.2478 ± 0.0011 mm3, n = 1150) than in H. yedoensis [0.3481 ± 0.0013 mm3, n = 1046; nested ANOVA, F 1,49 = 5458.62, P < 0.0001; female code (species): df = 49, F = 53.9136, P < 0.0001; clutch code (female code): df = 187, F = 6.3250, P < 0.0001]. There were significant maternal body size and species effects on ovariole number but no maternal body size × species interaction was detected (ANCOVA, maternal body size: F 1,47 = 9.09, P < 0.01; species: F 1,47 = 96.10, P < 0.0001; interaction: F 1,47 = 1.63, P = 0.21; Fig. 2).

Fig. 1
figure1

Relationship between female body size and egg size (mean ± SE) by species. Mean volumes of egg produced by H. axyridis (open circles) and H. yedoensis (closed circles) females of a given body length

Fig. 2
figure2

Relationship between female body size and number of ovarioles in H. axyridis (open circles) and H. yedoensis (closed circles). Dotted and solid lines represent the linear regression for H. axyridis and H. yedoensis, respectively

Model analysis

We present our results as graphical solutions owing to the complexity of the model. We first consider the special case in which the maternal resource in the trophic eggs is completely consumed by the offspring (i.e., δ = 1). The effects of the proportion of the good environment (p) on the optimal proportion of trophic eggs and on egg size are depicted graphically in Fig. 3. The optimal proportion of trophic eggs relative to p is a convex upward curve (Fig. 3a), and the optimal egg size decreases with p (Fig. 3b).

Fig. 3
figure3

Optimal proportion of trophic eggs (a) and optimal offspring size (b) as a function of the proportion of the good environment. Parameter values used are k = 2.0, q = 0.75 (solid line); k = 4.0, q = 0.75 (dashed line); and k = 2.0, q = 0.90 (dotted line); where k is the degree of difference between the two environments and q is environmental predictability

The effects of environmental predictability (q) on the optimal proportion of trophic eggs and on egg size are depicted graphically in Fig. 4. The optimal proportion of trophic eggs is always zero when it is not possible to predict the environment (q = 0.5), and it increases as predictability increases (Fig. 4a).

Fig. 4
figure4

Optimal proportion of trophic eggs (a) and optimal offspring size (b) as a function of environmental predictability. We did not evaluate the situation where q < 0.5 because it is biologically unrealistic. Parameter values used are k = 2.0, p = 0.50 (solid line); k = 4.0, p = 0.50 (dashed line); and k = 2.0, p = 0.25 (dotted line); where k is the difference between the two environments and p is the proportion of the good environment

The effects of the magnitude of the difference in quality between the good and poor environments (k) on the optimal proportion of trophic eggs and on egg size are depicted in Fig. 5. Both the proportion of trophic eggs and egg size increase with k, but the rate of increase in the proportion of trophic eggs decays as k increases (Fig. 5a), whereas optimal egg size increases linearly with k (Fig. 5b).

Fig. 5
figure5

Optimal proportion of trophic eggs (a) and optimal offspring size (b) as a function of the difference in quality between the good and bad environments. Parameter values used are p = 0.5, q = 0.66 (solid line); p = 0.25, q = 0.66 (dashed line); and p = 0.5, q = 0.75 (dotted line); where p is the proportion of the good environment and q is environmental predictability

Finally, we consider the case that the trophic egg resource is not completely consumed by the offspring (0 ≤ δ ≤ 1). Despite the costs associated with trophic egg consumption, a trophic egg strategy can evolve in heterogeneous environments (Fig. 6a). In fact, the greater the difference in quality between two environments, the larger the area in the parameter space where a trophic egg strategy is favoured (‘trophic egg area’, grey and black in Fig. 6). However, trophic egg area severely decreases as investment efficiency decrease, especially when the difference in quality between two environments is small (k = 1.5, Fig. 6b).

Fig. 6
figure6

Conditions that favour a trophic egg strategy (t > 0) when k = 1.5 (black area) or k = 2.0 (black and grey areas), depending on the investment efficiency δ (horizontal axis): a proportion of good environment p (vertical axis), and b environmental predictability q (vertical axis). Other parameters: a q = 0.75, b p = 0.25. In the shaded parameter area, a trophic egg-laying strategy is expected to never evolve

Discussion

Here we present three key findings in trophic egg evolution. First, intra- and interspecific morphological comparisons suggest that maternal body size is at best a weak morphological constraint on egg size in H. axyridis, which has been experimentally proved to produce trophic eggs (Perry and Roitberg 2005a), and H. yedoensis (Fig. 1). Second, we find that trophic egg-laying is expected to evolve in heterogeneous environments when mothers cannot manipulate egg size plastically (Fig. 3a). Third, we theoretically show that a small reduction in investment efficiency in tropic egg consumption greatly reduces the likelihood of trophic egg evolution, even when cue reliability is relatively high (Fig. 6). Taken together, these finding lead us to conclude that maternal adaptation to highly heterogeneous environments rather than morphological constraints on egg size is a sufficient explanation for the evolution of trophic egg-laying in some non-eusocial animals.

This is the first study to evaluate the role of morphological constraints on egg size in species that produce trophic eggs. We found no significant correlation between egg size and maternal body size in H. axyridis or H. yedoensis (Fig. 1), indicating that smaller females can produce eggs similar in size to the eggs of larger females. Moreover, the relationship between egg size and maternal body size was extremely weak in both species, accounting for <1 % of the total variation in egg size. Furthermore, H. yedoensis females produce larger eggs than H. axyridis females, despite the similar maternal body size in the two species (Fig. 1). A reduction in the number of ovarioles should contribute to the production of larger eggs relative to maternal body size (Fig. 2). This result is consistent with the previous finding that species of ladybirds with few ovarioles lay larger eggs than similar-sized species with many ovarioles (Stewart et al. 1991a). These results suggest that maternal body size as a morphological constraint has at best a minor role in the determination of egg size in the two studied ladybird species. Recent studies of insects also suggest that the importance of morphological constraints on the evolution of egg size has been overemphasized (Fischer et al. 2002; Bauerfeind and Fischer 2008; Noriyuki et al. 2010). Importantly, morphological constraints are particularly unlikely to exist when eggs are small relative to the size of the mother, as is the case in most insect species as well as tree frogs and sharks that produce large numbers of trophic eggs. In addition, even if morphological constraints prevent small females from producing eggs of optimal size, such females can overcome such constraints by producing more elongated eggs (Congdon and Gibbons 1987). In the subsocial bug Adomerus triguttulus, viable eggs are more elongated than trophic eggs (Kudo et al. 2006), suggesting that the females may have potential to overcome morphological constraints on egg size by changing the egg morphology. Moreover, in several reptiles, smaller mothers produce elongated eggs, presumably to facilitate their smooth passage out of the mother’s body (Sinervo and Licht 1991; Clark et al. 2001; Ji et al. 2006; Rollinson and Brooks 2008). Therefore, it is possible that morphological constraints may not adequately account for trophic egg evolution in other animals.

Instead, our theoretical model showed that trophic egg provisioning to small offspring is favoured in heterogeneous environments when mothers cannot manipulate egg size plastically (Fig. 3). Optimal per offspring maternal investment in a poor environment can also be achieved by the evolution of large eggs, without trophic egg-laying, because very large offspring can survive despite variation in environmental quality. As a result, in some circumstances selection favours females that consistently produce large eggs. Females following this non-plastic strategy, however, are obligated to invest an amount of resources in excess of the optimal value in the good environment, which does not require a large amount of maternal resources. In contrast, by following a trophic egg-laying strategy, females can change their per offspring maternal investment even after deposition of viable eggs, suggesting that trophic egg production and consumption by hatchlings allows females to flexibly adapt to a variable resource environment.

Note that the maximum proportion of trophic eggs is predicted when 0.5 < p < 1 (Fig. 3a). This result indicates that trophic egg-laying tends to be favoured when the proportion of the good environment is higher than the proportion of the poor environment. This result may appear counterintuitive, because we assume that mothers provision trophic eggs to deal with a poor environment. However, although the large egg strategy is inflexible, it can consistently achieve a high offspring survival rate even in a poor environment. By contrast, a trophic egg strategy may lead to large reductions of fitness and offspring survival in a poor environment if the mother incorrectly assesses the environmental quality and therefore fails to provide trophic eggs to small offspring. Thus, a large egg strategy, which should be a safe strategy even in a poor environment, may be favoured when the proportion of poor environment is relatively high.

Our conclusion that evolution of trophic eggs requires a highly heterogeneous environment is consistent with the empirical reports in both vertebrates and invertebrates (e.g., Crump 1981; Dixon 1998). For example, in aphidophagous lacewings and ladybirds, food resources are frequently and intermittently limited over time because of the ephemeral nature of aphid colonies (Osawa 1992b; Hemptinne et al. 1992; Dixon 1998), and they are also spatially heterogeneous in quality and quantity (Osawa 2000). In sub-social animals that provide parental care to offspring even after the hatching (e.g., tree frogs, burrower bugs, and passalid beetles), as well as predatory animals without effective natural enemies (e.g., sharks and ladybirds), mothers may have relatively long ecological longevity and thus may experience various environmental conditions over their reproductive period. Therefore, it is suggested that trophic egg provisioning may function as a flexible solution for dealing with multiple habitats (Perry and Roitberg 2006). Because in our model we do not distinguish between spatial and temporal variation, our findings are potentially applicable to diverse animal taxa producing trophic eggs to cope with predictable environmental variation.

Our model revealed that environmental predictability enhances the likelihood that trophic egg provisioning to small eggs will evolve (Fig. 4). Our result indicates that the evolution of trophic egg-laying is possible in heterogeneous environments if mothers have even a little ability to estimate the offspring’s environment (q ≥ 0.5). However, recall that this result holds only when the investment efficiency of trophic egg consumption δ is 1 (results with 0 ≤ δ ≤ 1 are discussed below).

In some species with trophic egg-laying, mothers are known to evaluate environment conditions that their offspring will face in several ways. For example, in tree frogs, mothers adjust the number of trophic eggs based on the number of offspring (Kam et al. 1998) and offspring age (Gibson and Buley 2004). In A. triguttulus females adjust the number of trophic eggs per viable egg in response to varying resource environments prior to oviposition (Kudo and Nakahira 2005). However, in a majority of trophic egg laying species, the role of environmental cues in trophic egg provisioning has not been examined (Perry and Roitberg 2006). In order to evaluate our model prediction that environmental predictability should be necessary for evolution of trophic egg-laying (Fig. 4), empirical tests are required to detect maternal plasticity in trophic egg-laying.

Our model also demonstrated that the evolution of trophic eggs is highly sensitive to the wasteful expenditure of maternal resource for trophic eggs even when environmental predictability is relatively high (Fig. 6). In particular, when environmental quality does not differ very much between the good and poor environments (k = 1.5 in Fig. 6), even a small reduction of investment efficiency makes the evolution of trophic egg-laying unlikely. Both nutritive parts of trophic eggs and the shells may be left uneaten by offspring (Perry and Roitberg 2005a). Moreover, non-sibling conspecific individuals and other predators attack trophic eggs, especially in species with no post-natal parental care but also in sub-social animals (Osawa 1989; Nomakuchi et al. 2001). We suggest therefore that these moderate but unavoidable costs associated with trophic egg consumption may mitigate against the evolution of trophic egg-laying, thus accounting for the evolution of trophic egg laying in some taxa and not others.

Variation in the division of maternal resources among siblings may impose both costs and benefits on maternal trophic egg provisioning, although our model does not explicitly consider this mechanism. Classical optimal investment theory predicts that the amount of parental resource per offspring should be constant in a given environment (Smith and Fretwell 1974). In real organisms, however, the amount of parental resource per offspring in a brood with trophic eggs frequently varies among siblings (Osawa 1992a), because hatching asynchrony and trophic egg location in the clutch, for example, may cause a bias in resource allocation (Osawa 1992a; Perry and Roitberg 2005a). Moreover, adults that abandon their eggs presumably have little ability to control the distribution of resources among offspring. Thus, it is suggested that biases in resource allocation among siblings may prevent mothers from producing trophic eggs. Conversely, trophic egg-laying may operate as a bet-hedging strategy by generating variation in the size of offspring, the largest of which can survive in the event of unpredictable poor food availability (Perry and Roitberg 2006). In support of this argument, some empirical and comparative studies suggest that within-clutch variation in egg size can reflect an adaptive strategy for dealing with in unpredictable environments in diverse animal taxa such as frogs and fishes (Crump 1981; Einum and Fleming 2004; Marshall et al. 2008; Crean and Marshall 2009). However, the bet-hedging hypothesis for trophic eggs has yet to be tested against the alternative hypothesis of a single optimum in provisioning per offspring. Further investigation of bet-hedging as an evolutionary mechanism promoting trophic egg provisioning should be a productive area of investigation.

Our model results are consistent with the findings of previous theoretical studies on adaptive phenotypic plasticity (Berrigan and Scheiner 2004). Specifically, models of phenotypic plasticity generally show that plasticity is favoured when (1) there is environmental heterogeneity (spatial or temporal), (2) there are cues that reliably predict future environmental conditions, and (3) the cost of plasticity is low. Therefore, we suggest that trophic egg provisioning can be regarded as one strategy of adaptive phenotypic plasticity when plastic adjustment of egg size is constrained.

In our model, parental-offspring conflict is not taken into account: We assumed that offspring cannot influence parental behaviour. Importantly, selection typically maximizes maternal rather than offspring fitness, particularly in species with no post-natal parental care, in which offspring counter-strategies may be less likely to evolve (Smith and Fretwell 1974; Trivers 1974). However, parental-offspring conflict affects the evolution of maternal reproductive strategies if offspring can counteract maternal strategies (Parker et al. 2002; Perry and Roitberg 2005b). In particular, Crespi (1992) discussed hypothetically the evolution of trophic eggs in the context of reduction of parent-offspring conflict over sibling cannibalism. Specifically, he suggests that when parent and offspring interests conflict over sibling cannibalism, mothers might adopt a strategy to limit cannibalism by producing trophic eggs, which are less costly than viable eggs but which provide enough energy to cause offspring to refrain from eating viable siblings (Crespi 1992). This hypothesis and our predictions concerning environmental heterogeneity are not necessarily mutually exclusive. Models that incorporate the offspring’s point of view need to be developed to predict whether Crespi’s (1992) argument can function as a general explanation for trophic egg evolution. However, some empirical data refute the generality of the parent-offspring conflict reduction hypothesis. In particular, Kudo and Nakahira (2004) explicitly rejected the hypothesis by showing in careful experiments in the sub-social burrower bug that the presence or absence of trophic eggs did not affect the rate of sibling cannibalism. Moreover, a cost difference between trophic and viable egg production may not be common in animal species (Perry and Roitberg 2006). Nevertheless, further studies are needed to identify possible differences in quality between trophic and viable eggs to evaluate the parent-offspring conflict reduction hypothesis.

References

  1. Alexander RD (1974) The evolution of social behavior. Annu Rev Ecol Syst 5:325–383

    Article  Google Scholar 

  2. Bauerfeind SS, Fischer K (2008) Maternal body size as a morphological constraint on egg size and fecundity in butterflies. Basic Appl Ecol 9:443–451

    Article  Google Scholar 

  3. Baur B (1988) Repeated mating and female fecundity in the simultaneously hermaphroditic land snail Arianta arbustorum. Int J Invertebr Reprod Dev 14:197–204

    Article  Google Scholar 

  4. Baur B (1990) Possible benefits of egg cannibalism in the land snail Arianta arbustorum (L.). Funct Ecol 4:679–684

    Article  Google Scholar 

  5. Baur B, Raboud C (1988) Life history of the land snail Arianta arbustorum along an altitudinal gradient. J Anim Ecol 57:71–87

    Article  Google Scholar 

  6. Berrigan D, Scheiner SM (2004) Modeling the evolution of phenotypic plasticity. In: DeWitt TJ, Scheiner SM (eds) Phenotypic plasticity: functional and conceptual approaches. Oxford University Press, Oxford, pp 82–97

    Google Scholar 

  7. Christians JK (2002) Avian egg size: variation within species and inflexibility within individuals. Biol Rev 77:1–26

    PubMed  Article  Google Scholar 

  8. Clark PJ, Ewert MA, Nelson CE (2001) Physical apertures as constraints on egg size and shape in the common musk turtle, Sternotherus odoratus. Funct Ecol 15:70–77

    Article  Google Scholar 

  9. Congdon JD, Gibbons JW (1987) Morphological constraint on egg size: a challenge to optimal egg size theory? Proc Natl Acad Sci USA 84:4145–4147

    PubMed  Article  CAS  Google Scholar 

  10. Crean AJ, Marshall DJ (2009) Coping with environmental uncertainty: dynamic bet hedging as a maternal effect. Philos Trans R Soc B 364:1087–1096

    Article  Google Scholar 

  11. Crespi BJ (1992) Cannibalism and trophic egg in subsocial and eusocial insects. In: Elgar MA, Crespi BJ (eds) Cannibalism: ecology and evolution among diverse taxa. Oxford University Press, Oxford, pp 176–213

    Google Scholar 

  12. Crump ML (1981) Variation in propagule size as a function of environmental uncertainty for tree frogs. Am Nat 117:724–737

    Article  Google Scholar 

  13. Dall SRX, Giraldeau LA, Olsson O, McNamara JM, Stephens DW (2005) Information and its use by animals in evolutionary ecology. Trends Ecol Evol 20:187–193

    PubMed  Article  Google Scholar 

  14. Dixon AFG (1998) Aphid ecology. Chapman and Hall, London

    Google Scholar 

  15. Dixon AFG (2000) Insect predator–prey dynamics: ladybird beetles and biological control. Cambridge University Press, Cambridge

    Google Scholar 

  16. Dixon AFG, Guo Y (1993) Egg and cluster size in ladybird beetles (Coleoptera: Coccinellidae): the direct and indirect effects of aphid abundance. Eur J Entomol 90:457–463

    Google Scholar 

  17. Einum S, Fleming IA (2004) Environmental unpredictability and offspring size: conservative versus diversified bet-hedging. Evol Ecol Res 6:443–455

    Google Scholar 

  18. Elgar MA, Crespi BJ (1992) Cannibalism: ecology and evolution among diverse taxa. Oxford University Press, Oxford

    Google Scholar 

  19. Fischer K, Zwaan BJ, Brakefield PM (2002) How does egg size relate to body size in butterflies? Oecologia 131:375–379

    Article  Google Scholar 

  20. Fischer B, Taborsky B, Kokko H (2011) How to balance the offspring quality–quantity tradeoff when environmental cues are unreliable. Oikos 120:258–270

    Article  Google Scholar 

  21. Fox CW, Czesak ME (2000) Evolutionary ecology of progeny size in arthropods. Annu Rev Entomol 45:341–369

    PubMed  Article  CAS  Google Scholar 

  22. Fox CW, Thakar MS, Mousseau TA (1997) Egg size plasticity in a seed beetle: an adaptive maternal effect. Am Nat 149:149–163

    Article  Google Scholar 

  23. Gibson RC, Buley KR (2004) Maternal care and obligatory oophagy in Leptodactylus fallax: a new reproductive mode in frogs. Copeia 2004:128–135

    Article  Google Scholar 

  24. Gilbert F (1990) Size, phylogeny and life-history in the evolution of feeding specialization in insect predators. In: Gilbert F (ed) Insect life cycles: genetics, evolution and co-ordination. Springer, New York, pp 101–124

    Google Scholar 

  25. Gobin B, Peeters C, Billen J (1998) Production of trophic eggs by virgin workers in the ponerine ant Gnamptogenys menadensis. Physiol Entomol 23:329–336

    Article  Google Scholar 

  26. Hemptinne JL, Dixon AFG, Coffin J (1992) Attack strategy of ladybird beetles (Coccinellidae): factors shaping their numerical response. Oecologia 90:238–245

    Google Scholar 

  27. Hölldobler B, Wilson EO (1990) The ants. Harvard University Press, Massachusetts

    Google Scholar 

  28. Ji X, Du WG, Li H, Lin LH (2006) Experimentally reducing clutch size reveals a fixed upper limit to egg size in snakes, evidence from the king ratsnake, Elaphe carinata. Comp Biochem Physiol A 144:474–478

    Article  Google Scholar 

  29. Kam YC, Lin CF, Lin YS, Tsal YF (1998) Density effects of oophagous tadpoles of Chirixalus eiffingeri (Anura: Rhacophoridae): importance of maternal brood care. Herpetologica 54:425–433

    Google Scholar 

  30. Kawai A (1978) Sibling cannibalism in the first instar larvae of Harmonia axyridis Pallas (Coleoptera: Coccinelidae). Kontyû 46:14–19

    Google Scholar 

  31. Kawecki TJ (1995) Adaptive plasticity of egg size in response to competition in the cowpea weevil, Callosobruchus maculatus (Coleoptera: Bruchidae). Oecologia 102:81–85

    Google Scholar 

  32. Kudo S, Nakahira T (2004) Effects of trophic-eggs on offspring performance and rivalry in a sub-social bug. Oikos 107:28–35

    Article  Google Scholar 

  33. Kudo S, Nakahira T (2005) Trophic-egg production in a subsocial bug: adaptive plasticity in response to resource conditions. Oikos 111:459–464

    Article  Google Scholar 

  34. Kudo S, Nakahira T, Saito Y (2006) Morphology of trophic eggs and ovarian dynamics in the subsocial bug Adomerus triguttulus (Heteroptera: Cydnidae). Can J Zool 84:723–728

    Article  Google Scholar 

  35. Leather SR, Burnand AC (1987) Factors affecting life-history parameters of the pine beauty moth, Panolis flammea (D & S): the hidden costs of reproduction. Funct Ecol 1:331–338

    Article  Google Scholar 

  36. Majerus TMO, Majerus MEN, Knowles B, Wheeler J, Bertrand D, Kuznetzov VN, Ueno H, Hurst GDD (1998) Extreme variation in the prevalence of inherited male-killing microorganisms between three populations of Harmonia axyridis (Coleoptera: Coccinellidae). Heredity 81:683–691

    Article  Google Scholar 

  37. Marshall DJ, Uller T (2007) When is a maternal effect adaptive? Oikos 116:1957–1963

    Article  Google Scholar 

  38. Marshall DJ, Bonduriansky R, Bussière LF (2008) Offspring size variation within broods as a bet-hedging strategy in unpredictable environments. Ecology 89:2506–2517

    PubMed  Article  Google Scholar 

  39. McGinley MA, Temme DH, Geber MA (1987) Parental investment in offspring in variable environments: theoretical and empirical considerations. Am Nat 130:370–398

    Article  Google Scholar 

  40. Mizumoto M, Nakasuji F (2007) Egg size manipulation in the migrant skipper Parnara guttata guttata (Lepdoptera: Hesperiidae), in response to different host plants. Popul Ecol 49:135–140

    Article  Google Scholar 

  41. Mock DW, Parker GA (1997) The evolution of sibling rivalry. Oxford University Press, Oxford

    Google Scholar 

  42. Nomakuchi S, Filippi L, Hironaka M (2001) Nymphal occurrence pattern and predation risk in the subsocial shield bug, Parastrachia japonensis (Heteroptera: Cydnidae). Appl Entomol Zool 36:209–212

    Article  Google Scholar 

  43. Noriyuki S, Kishi S, Nishida T (2010) Seasonal variation of egg size and shape in Ypthima multistriata (Lepidoptera: Satyridae) in relation to maternal body size as a morphological constraint. Ann Entomol Soc Am 103:580–584

    Article  Google Scholar 

  44. Noriyuki S, Osawa N, Nishida T (2011) Prey capture performance in hatchlings of two sibling Harmonia ladybird species in relation to maternal investment through sibling cannibalism. Ecol Entomol 36:282–289

    Article  Google Scholar 

  45. Osawa N (1989) Sibling and non-sibling cannibalism by larvae of a lady beetle Harmonia axyridis Pallas (Coleoptera: Coccinellidae) in the field. Res Popul Ecol 31:153–160

    Article  Google Scholar 

  46. Osawa N (1992a) Sibling cannibalism in the ladybird beetle Harmonia axyridis Pallas: fitness consequences for mothers and offspring. Res Popul Ecol 34:45–55

    Article  Google Scholar 

  47. Osawa N (1992b) A life table of the ladybird beetle Harmonia axyridis Pallas (Coleoptera, Coccinellidae) in relation to the aphid abundance. Jap J Entomol 60:575–579

    Google Scholar 

  48. Osawa N (2000) Population field studies on the aphidophagous ladybird beetle Harmonia axyridis (Coleoptera: Coccinellidae): resource tracking and population characteristics. Popul Ecol 42:115–127

    Article  Google Scholar 

  49. Osawa N, Ohashi K (2008) Sympatric coexistence of sibling species Harmonia yedoensis and H. axyridis (Coleoptera: Coccinellidae) and the roles of maternal investment through egg and sibling cannibalism. Eur J Entomol 105:445–454

    Google Scholar 

  50. Osawa N, Yoshinaga A (2009) The presence of micropyles in the shells of developing and undeveloped eggs of the ladybird beetle Harmonia axyridis (Coleoptera: Coccinellidae). Eur J Entomol 106:607–610

    Google Scholar 

  51. Parker GA, Begon M (1986) Optimal egg size and clutch size: effects of environment and maternal phenotype. Am Nat 128:573–592

    Article  Google Scholar 

  52. Parker GA, Royle NJ, Hartley IR (2002) Intrafamilial conflict and parental investment: a synthesis. Philos Trans R Soc B 357:295–307

    Article  Google Scholar 

  53. Perry JC, Roitberg BD (2005a) Ladybird mothers mitigate offspring starvation risk by laying trophic eggs. Behav Ecol Sociobiol 58:578–586

    Article  Google Scholar 

  54. Perry JC, Roitberg BD (2005b) Games among cannibals: competition to cannibalize and parent-offspring conflict lead to increased sibling cannibalism. J Evol Biol 18:1523–1533

    PubMed  Article  CAS  Google Scholar 

  55. Perry JC, Roitberg BD (2006) Trophic egg laying: hypotheses and tests. Oikos 112:706–714

    Article  Google Scholar 

  56. Polis GA (1984) Intraspecific predation and “infant killing” among invertebrates. In: Hausfater G, Hrdy SB (eds) Infanticide: comparative and evolutionary perspectives. Aldine De Gruyter, Berlin, pp 87–104

    Google Scholar 

  57. Rollinson N, Brooks R (2008) Optimal offspring provisioning when egg size is “constrained”: a case study with the painted turtle Chrysemys picta. Oikos 117:144–151

    Article  Google Scholar 

  58. Sakagami SF (1982) Stingless bees. In: Hermann HR (ed) Social insects, vol 3. Academic Press, Massachusetts, pp 362–424

    Google Scholar 

  59. Sakai S, Harada Y (2001) Why do large mothers produce large offspring? Theory and a test. Am Nat 157:348–359

    PubMed  Article  CAS  Google Scholar 

  60. Sasaji H (1998) Natural history of the ladybirds. University of Tokyo Press, Tokyo (in Japanese)

    Google Scholar 

  61. Sinervo B, Licht P (1991) Hormonal and physiological control of clutch size, egg size, and egg shape in side-blotched lizards (Uta stansburiana): constraints on the evolution of lizard life histories. J Exp Zool 257:252–264

    Article  CAS  Google Scholar 

  62. Smith CC, Fretwell SD (1974) The optimal balance between size and number of offspring. Am Nat 108:499–506

    Article  Google Scholar 

  63. Soares AO, Coderre D, Schanderl H (2001) Fitness of two phenotypes of Harmonia axyridis (Coleoptera: Coccinellidae). Eur J Entomol 98:287–293

    Google Scholar 

  64. Sokal RR, Rohlf FJ (1995) Biometry, 3rd edn. Freeman, New York

    Google Scholar 

  65. Stewart LA, Hemptinne JL, Dixon AFG (1991a) Reproductive tactics of ladybird beetles: relationship between egg size, ovariole number and developmental time. Funct Ecol 5:380–385

    Article  Google Scholar 

  66. Stewart LA, Dixon AFG, Ruzicka Z, Iperti G (1991b) Clutch and egg size in ladybird beetles. Entomophaga 36:93–97

    Article  Google Scholar 

  67. Takakura K (2004) Variation in egg size within and among generations of the bean weevil, Bruchidius dorsalis (Coleoptera, Bruchidae): effects of host plant quality and paternal nutritional investment. Ann Entomol Soc Am 97:346–352

    Article  Google Scholar 

  68. Trivers RL (1974) Parent-offspring conflict. Am Zool 14:249–264

    Google Scholar 

  69. Wedell N, Gage MJG, Parker GA (2002) Sperm competition, male prudence and sperm-limited females. Trends Ecol Evol 17:313–320

    Article  Google Scholar 

  70. Werner EE, Gilliam JF (1984) The ontogenetic niche and species interactions in size structures populations. Annu Rev Ecol Syst 15:393–425

    Article  Google Scholar 

Download references

Acknowledgments

We thank T. Nishida and N. Baba for valuable discussions, Y. Harada for advice on mathematical analyses, S. Seiter for improving English version, and M. Tokeshi and J.-Y. Ide for comments on the manuscript. We are grateful to the staff at the Botanical Garden of Kyoto University for permission to collect ladybirds. This study was supported by a Research Fellowship for Young Scientists from the Japan Society for the Promotion of Science to S. Noriyuki and K. Kawatsu, and a Grant-in-Aid for Scientific Research (Ministry of Education, Culture, Sports, Science and Technology of Japan; No. 20405047) to N. Osawa.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Suzuki Noriyuki.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Noriyuki, S., Kawatsu, K. & Osawa, N. Factors promoting maternal trophic egg provisioning in non-eusocial animals. Popul Ecol 54, 455–465 (2012). https://doi.org/10.1007/s10144-012-0317-6

Download citation

Keywords

  • Egg size
  • Environmental heterogeneity
  • Harmonia
  • Maternal investment
  • Morphological constraint
  • Phenotypic plasticity