Behavioral Ecology Advance Access originally published online on May 11, 2005
Behavioral Ecology 2005 16(4):779-787; doi:10.1093/beheco/ari053
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Egg size and reproductive allocation in eusocial thrips
School of Biological Sciences, Flinders University, Bedford Park, SA 5042, Australia; School of Botany and Zoology, Australian National University, Canberra, ACT 0200, Australia; and School of Earth and Environmental Sciences, University of Adelaide, Adelaide, SA 5005, Australia
Address correspondence to B.D. Kranz. E-mail: brenda.kranz{at}adelaide.edu.au.
Received 30 September 2004; revised 9 March 2005; accepted 7 April 2005.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Reproductive allocation, in terms of fecundity and egg size, has been given little consideration in eusocial societies. To begin to address this, absolute and body size–adjusted egg volumes were compared, along with fecundity, between the foundress and her subfertile soldier offspring in the eusocial, gall-inducing thrips, Kladothrips hamiltoni, Kladothrips waterhousei, and Kladothrips habrus, and a congeneric, Kladothrips morrisi, with fully fecund soldiers. Soldiers produced significantly larger eggs than the foundress in all species except K. morrisi, where egg volumes did not differ. After accounting for body size, soldiers produced significantly smaller eggs than the foundress in K. morrisi and marginally so in K. waterhousei, but egg sizes did not differ in K. hamiltoni and K. habrus. When egg size and fecundity data are combined, K. morrisi soldiers invest less in reproduction than the foundress, and in conjunction with other life-history features the species can be considered eusocial. Maximum likelihood analyses reveal relatively low reproductive allocation skew in the ancestral lineages and high skew in the derived lineages, but the trend is not significant when fecundity and egg size are considered separately. Gall size covaried negatively with soldier-to-foundress relative body size–adjusted egg size and reproductive allocation and marginally so with fecundity, suggesting that gall size is a determinant of egg size and fecundity trade-offs in eusocial thrips and providing the strongest support to date that gall size has featured in the social evolution of this clade. This study highlights that data on fecundity alone may be insufficient for assessing reproductive division of labor.
Key words: egg size, eusociality, density-dependent selection, galling thrips, phylogenetic directionality, reproductive allocation.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Reproductive allocation and the relationship between offspring size and number (Smith and Fretwell, 1974
) are major components of maternal investment, parental care, and fitness (Clutton-Brock, 1991; McEdward and Morgan, 2001; Sargent et al., 1987; Shine, 1978). Assuming a fixed investment potential and positive relationship between offspring fitness and maternal investment per progeny (offspring size), the optimal reproductive allocation theory predicts a trade-off between fecundity and offspring size to optimize maternal fitness (Smith and Fretwell, 1974; Parker and Begon, 1986). Such a trade-off has been identified in a range of solitary insects, crustaceans, and other invertebrates (references in Brown, 2003; Fox and Czesak, 2000), as well as voles (Mappes and Koskela, 2004) and fish (Heins et al., 2004), although other studies have shown a lack of correlation (Fox and Czesak, 2000; Roff, 2002).
Among the many traits associated with offspring size in animals (Clutton-Brock, 1991; Fox and Czesak, 2000), relatively large offspring can lead to early maturation (Laptikhovskii, 2002; Mappes and Koskela, 2004; Santo et al., 2001; but see Valdimarsson et al., 2002) and better competitive capabilities, especially as juveniles (McLain and Mallard, 1991; Svensson and Sinervo, 2000). It could be particularly useful to produce large offspring with such qualities in late broods, the mechanism of which might be explained by two alternative hypotheses. First, a female with reduced reproductive value late in life might increase her fitness by producing larger offspring as a form of "terminal investment" (Clutton-Brock, 1984). Such a strategy has been identified in a range of animals, including the redback spider (Andrade and Banta, 2002), house sparrow (Bonneaud et al., 2004), and moth (Javois and Tammaru, 2004). Second, late brood may be subject to higher levels of density-dependent competition than their earlier born siblings, and faster maturation and a competitive edge would be especially useful under these conditions (Charlesworth, 1971; Clarke, 1972; Smith and Fretwell, 1974; Winkler and Wallin, 1987; Witting, 1997). Several studies have shown that egg size is positively associated with clutch size, where large eggs under high-density conditions increase the competitive success and maturation rate of the offspring relative to their earlier born siblings (Brockelman, 1975; Sinervo and Doughty, 1996; Svensson and Sinervo, 2000).
Reproductive allocation has been investigated in some cooperative bird studies (Liker et al., 2001; Vehrencamp, 2000), but such studies are lacking in eusocial groups. In eusocial societies, fecundity is skewed in favor of a dominant individual or individuals (Michener, 1969; Wilson, 1971), and yet there is no information regarding egg size–related investment between reproductive castes, despite this being as directly quantifiable as fecundity. Determining egg size variation between castes in eusocial species will lead to a more concise evaluation of the level of reproductive division of labor between castes and provide a more comprehensive understanding of the strategies employed in maternal investment where there may be conflict over reproduction.
This paper investigates intercaste egg size in the foundress (dominant reproductive) and soldiers (subordinates) of eusocial, haplodiploid, gall-inducing thrips and combines the data with that of relative fecundity from Kranz et al. (1999, 2001a,b) to compare total reproductive allocation within and across species. Such data will provide a more complete picture of reproductive dynamics between castes within species and can be used for phylogenetic contrasts of reproduction between species.
Eusocial, gall-inducing thrips
Of the 21 gall-inducing thrips species described from Australian Acacia, seven species in the genus Kladothrips have a nondispersing soldier caste (Crespi et al., 2004; Wills et al., 2004; Figure 1). Soldiers of both sexes comprise the first-generation offspring of a single gall-inducing foundress. Relative to the foundress, soldiers have an enlarged prothorax and fore femora, reduced wings and antennae, and a pale exoskeleton (Crespi and Mound, 1997; Crespi et al., 2004; Mound, 1971; Mound et al., 1996). Behavioral observations in Kladothrips hamiltoni, Kladothrips morrisi (formerly Oncothrips morrisi), Kladothrips waterhousei (formerly Oncothrips waterhousei), Kladothrips habrus (formerly Oncothrips habrus), and Kladothrips intermedius (formerly Oncothrips tepperi) reveal that soldiers fight and often kill invading kleptoparasitic Koptothrips, hymenopterans, and lepidopterans (Crespi, 1992; Mound and Crespi, 1995; Perry et al., 2004) and that species in the derived lineages are more inclined and effective in defending against Koptothrips than those in the more basal lineages (Perry et al., 2003, 2004).
|
There appears to be a relationship between gall size and fecundity in the soldier-producing species where, in general, the more derived species have the smallest galls and soldiers have the lowest fecundity, both absolutely and relative to the foundress (Kranz et al., 1999
, 2001a,b; Wills et al., 2001; Figure 1). Based on the apparent, but statistically untested, phylogenetic directionality of reproductive division of labor in thrips, it has been speculated that selection for defense occurred before alloparental care and subfertility had evolved (Chapman et al., 2002; Crespi et al., 2004; Kranz et al., 2001b).
Soldiers sibmate and, along with the foundress, produce offspring of both sexes that disperse and become the next generation of gall foundresses and their mates (Crespi, 1992; Crespi et al., 1997, 2004; Kranz et al., 1999, 2001a,b; Mound and Crespi, 1995). The foundress dies in the gall and the soldiers continue to reproduce. While the soldiers produce less disperser offspring than the foundress on a per capita basis, soldiers in most species collectively produce the great majority of the dispersers, who are also the last brood in the gall (Chapman et al., 2002; Crespi, 1992; Crespi et al., 2004; Kranz et al., 1999, 2001a,b; Figure 1). Consequently, most offspring of the soldiers are at a much higher density, with diminishing space for egg laying and feeding, compared to those of the foundress.
Egg laying within a gall can be divided into three main stages: (1) prior to first soldier eclosion and reproduction, the foundress produces eggs that become nondispersing soldiers and begins to lay eggs that become dispersers; (2) soldiers eclose and become reproductive, and the foundress and soldiers simultaneously produce eggs that become dispersers; and (3) the foundress dies, and soldiers continue to produce eggs that become dispersers. The maximum number of soldiers in a gall has been determined elsewhere for the species being considered here (Kranz et al., 1999, 2001a,b), and so it is possible to exclude galls with foundress-produced eggs that are destined to become soldiers in order to only consider eggs that are destined to become dispersers. Therefore, comparisons can be made between the sizes of disperser-destined eggs produced by the foundress and soldiers and can be used to investigate reproductive dynamics when the foundress and soldiers are together in a gall. Combined with data on relative fecundity of the foundress and soldiers (Kranz et al., 1999, 2001a,b), egg to body size ratios can produce an informative measure of reproductive allocation and a more complete picture of reproductive division of labor than by considering fecundity alone.
The primary aim of this study was to elucidate reproductive allocation patterns in thrips, both between the foundress and soldiers within a species and in a phylogenetic context between species, to further our understanding of the selective mechanisms for the evolution of eusociality in these taxa. The main hypothesis being addressed is that disperser-destined eggs produced by soldiers will be larger than those produced by the foundress, both absolutely and after adjusting for body size. Results are considered in the contexts of reproductive allocation and division of labor, density-dependent selection and terminal investment, foundress-soldier conflict, and phylogenetic directionality.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Study species and collections
The volumes of foundress- and soldier-produced eggs were determined for four thrips species with a nondispersing soldier caste: K. hamiltoni, K. morrisi, K. waterhousei, and K. habrus (Table 1, Figure 1).
|
K. hamiltoni induces galls on Acacia cambagei (Kranz et al., 1999
; Mound and Crespi, 1995) and is in the most basal lineage of the eusocial clade (Morris et al., 2001; Figure 1). Soldiers are protandrous, and per capita and total soldier reproduction is high relative to species in the more derived lineages (Kranz et al., 1999; Figure 1). Galls used in this study were collected from two sites along Arckaringa Creek, approximately 1100 km north of Adelaide, South Australia (Kranz et al., 1999; Table 1).
K. morrisi induces galls on Acacia calcicola (Kranz et al., 2001b; Mound et al., 1996) and is in the next most basal lineage (Morris et al., 2001; Figure 1). Unlike other species in the clade, there is no difference in per capita disperser offspring production between the foundress and female "soldiers," and so the term "nondispersing fighting morph" rather than soldier has been coined for this species (Kranz et al., 2001b). However, nondispersing morphs of all species considered here will be referred to as soldiers. The parity in reproduction could be partly due to the relatively large galls, reducing competition between female soldiers for feeding and egg-laying sites (Kranz et al., 2001b; Wills et al., 2001). There is no sex-specific order of soldier eclosion (Kranz et al., 2001b; Figure 1). Galls used in this study were collected 50 km N of Coober Pedy, along the Stuart Highway (Table 1).
K. waterhousei and K. habrus are very similar in morphology and life history (Kranz et al., 2001a; Mound et al., 1996), although they induce galls on different Acacia hosts. Soldiers are protogynous in both species, but their per capita and total reproduction relative to the foundress is lower in K. habrus than in K. waterhousei (Kranz et al., 2001a; Figure 1). K. waterhousei galls were collected from Acacia papyrocarpa at Middleback pastoral station, 20 km SW of Whyalla, South Australia, and K. habrus galls were collected from Acacia melvillei, 21 km SW of Mildura, Victoria (Table 1).
Galls of each species were part of collections for the life-history studies of Kranz et al. (1999, 2001a,b). Collections were transported on ice to Flinders University, where census data were obtained from intact galls without kleptoparasites. Eggs were measured, as below, in some of these galls (Table 1). For each species, galls where the only adult was a gall foundress and the number of larvae and pupae was less than the observed maximum number of soldiers in a gall (K. hamiltoni, 59; K. morrisi, 81; K. waterhousei, 18; K. habrus, 15; Kranz et al., 1999, 2001a,b) were excluded to remove the confounding effect of foundress-produced eggs destined to become soldiers.
Egg size and adult body measurements
During the census of gall contents under a Leica dissecting microscope, 10 randomly chosen eggs from each gall were mounted onto a slide in a 1:1 v/v solution of 0.9% NaCl and glycerol. All eggs were mounted from galls with less than 10 eggs. The length and equatorial width of each egg was measured to ±2.5 µm at 400x magnification under an Olympus compound microscope. The volume of each egg was approximated using the formula for the volume of a prolate spheroid, (4/3)
a2c (Beyer, 1987), where a is the equatorial radius (half the width) and c is the polar radius (length).
Adult body size needs to be accounted for when comparing egg sizes and assessing reproductive allocation. For each gall with either just a foundress (soldiers yet to eclose) or just soldiers (a dead foundress), the following structures were measured in the foundress or two randomly chosen female soldiers: head length (hl), head width (hw) just posterior to the eyes, medial pronotum length (pl), and pronotum width (pw) at the anterior and posterior edges, from which the mean width was used. Pronotum and head surface areas were chosen to represent structures that are and are not associated with fighting specialization, respectively (i.e., are and are not allometric between the foundress and soldiers). Body length was not considered an appropriate measure as it is telescopic and varies substantially according to reproductive status and diet (Kranz BD, unpublished data). The index was derived as (hl x hw) + (pl x pw). Using this index, a mean egg to body size ratio was determined for each gall.
Data analyses
Data were analyzed using SPSS® version 12.0.2. The validity of using head and pronotum lengths and widths to produce a body size index was assessed using Principal component analysis (PCA). Data for each species and gall type (adults: foundress only; foundress and female soldiers; soldiers only) were tested by site and collection date with Levene's test for homoscedasticity and nested ANOVAs (gall nested within site or month), with "gall" as a random factor. Data were log10 transformed where necessary to establish homoscedasticity. Nested ANOVAs (eggs within gall) further tested egg volume data between gall types for each species. Partial correlations of total brood size (all juveniles and, if present, soldiers) and egg volume, accounting for adult body size, collection date, and site, were performed for each species using the gall types foundress only and soldiers only. To compare reproductive investment in an offspring between the foundress and soldiers, the regression residuals and/or and ratios of egg volume and body volume were compared between morphs for each species using independent t tests. Body size–adjusted egg size and relative fecundity of the soldiers relative to the foundress were combined to give a measure of total reproductive allocation.
The software program Continuous© (Pagel, 1997, 1999) used maximum likelihood analysis across species to assess whether soldier-to-foundress average (1) fecundity, (2) egg size, (3) body size–adjusted egg size, and (4) reproductive allocation have phylogenetic directionality. Species-specific average gall size was also assessed. The analyses had a null hypothesis of random drift ("Model A") and an alternative hypothesis of directional section ("Model B") and used the tree branch lengths derived from Morris et al. (2001). Maximum likelihood analyses also tested for covariance across species of the same reproductive parameters, with gall size accounting for any phylogenetic dependence between species.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Pooling data by site and collection date
Egg volume data were assessed for site and collection date effects within each gall type and species (Table 1). Some sites and months were excluded due to heteroscedasticity, even after log10, ln, and square root transformations (data not shown). Of the raw data further analyzed, all were homoscedastic for all species and gall types except for K. morrisi galls with a foundress and soldiers together, and these data were homoscedastic after log10 transformation. For all gall types in all species, there was a highly significant effect of gall (p
.006), that is, there was variation in egg volume with family. After accounting for the gall effect, there was no effect of collection date or site for any of the gall types (Table 1). Consequently, egg volume data were pooled by gall type and species for further analyses.
Egg volume and brood size
It was important to investigate whether egg volume varied with total brood size within the gall types foundress only and soldiers only, as such a relationship could be biologically meaningful and a confounding effect for assessing differences in egg volumes between the foundress and soldiers. As body size positively correlated with egg volume in K. hamiltoni and K. morrisi (Spearman rank, K. hamiltoni foundresses: p < .001, soldiers: p = .029; K. morrisi foundresses p = .002, soldiers: p < .001; K. waterhousei foundresses: p = .312, soldiers: p = .484; K. habrus foundresses: p = .433, soldiers: p = .200) and there was only one body size datum per gall, the average egg volume for each gall was used in the analysis and, consequently, gall was not used as a covariate. There was no relationship between brood size and egg volume for either the foundress or soldiers in any of the species after accounting for body size and collection date (and site for K. hamiltoni soldier-only galls) (partial correlations, K. hamiltoni foundress: r5 = .68, p = .092; soldiers: r8 = –.04, p = .891; K. morrisi foundress: r8 =.14, p = .692; soldiers: r14 = –.32, p = .234; K. waterhousei foundress r7 = .21, p =.597; soldiers: r8 = –.01, p = .986; K. habrus r7 = –.25, p = .512; soldiers: r6 = –.08, p = .853).
Adult morph and egg volume
Foundress-produced eggs destined to be soldiers could not be compared to those destined to become dispersers as there were insufficient or no galls with eggs just destined to become soldiers. Species-specific frequency distributions for egg volumes combining all gall types revealed that the data were basically unimodal and that the volumes of foundress- and soldier-produced eggs partly overlapped (data not shown). Foundress-only galls had significantly smaller eggs than soldier-only galls for K. hamiltoni, K. waterhousei, and K. habrus, but there was no significant difference in such egg volumes for K. morrisi. Eggs in the gall type foundress and female soldiers were significantly larger than those of soldiers only in K. hamiltoni, significantly larger than those of foundress only but smaller than those of soldiers only in K. waterhousei, no different to those of foundress only and significantly smaller than those of soldiers only in K. habrus, and not significantly different to those of foundress only and soldiers only in K. morrisi (Table 2, Figure 2).
|
|
Reproductive allocation
Factor 1 of a PCA for head length, head width, pronotum length, and pronotum width ranged from 49% in K. waterhousei to 89% in K. morrisi, and communalities for all structures in all species ranged from 0.93 to 1, indicating that these pooled measurements provide a reasonable index of body size. Egg volume to body size ratios were compared between the gall types foundress only and soldiers only for each species to assess per capita egg investment of the foundress and soldiers. Comparing the unstandardized residuals of egg volume to body size may be a more powerful analysis than comparing just the ratios, but residuals could only be assessed for K. hamiltoni and K. morrisi as there was no relationship between these two variables for the other species (see above). Consequently, the ratio was compared between morphs for all species, and for K. hamiltoni and K. morrisi the result was compared to that of the residuals.
There was no significant difference in mean body size–adjusted egg volumes between the foundress and soldiers for K. hamiltoni (residuals: t22 = 1.70, p = .101; ratios: t22 = 0.59, p = .560) and K. habrus (ratios: t19 = –1.04, p = .312) (Figure 3). For K. morrisi, soldiers had significantly smaller body size–adjusted egg volumes than the foundress (assuming unequal variances, residuals: t17.1 = 4.04, p = .001; ratios: t14.2 = 4.72, p = .001), and the mean egg volume to body index ratio of the soldiers was 27% smaller than that of the foundress (Figure 3). For K. waterhousei, there was no significant difference in the egg volume to body size ratio between the foundress and soldiers at 
= 0.05 (t21 = 1.91, p = .070), but at = 0.1 the mean ratio of the soldiers was 12% smaller.
|
When considered with per capita disperser production (relative fecundity; Figure 1), the data indicate that the reproductive allocation (egg size and fecundity) of soldiers in K. morrisi is about 83% that of the foundress, which contrasts with fecundity at about 110% that of the foundress but not significantly different to parity. If the body size–adjusted egg volume data of the foundress and soldiers are considered statistically different in K. waterhousei then, using the same approach as for K. morrisi, the total per capita reproductive allocation of soldiers relative to the foundress is reduced from about 47%, for fecundity alone, to 35% when egg volume is incorporated. The reproductive allocation of soldiers to foundress in K. hamiltoni and K. habrus does not change from that of the fecundity data, at 47% and 15%, respectively (Figure 1).
Phylogenetic directionality and covariance with gall size
Phylogenetic directionality was assessed across species for per capita soldier-to-foundress relative fecundity (Figure 1), egg size (derived from Table 2), body size–adjusted egg volume and reproductive allocation, as well as gall size (Figure 1). The fecundity analysis also included data for K. intermedius as it is in the most derived lineage (Kranz et al., 2001b; Figure 1). Consequently, the reproductive allocation analysis was performed with and without the fecundity-only data of K. intermedius in order to account for and compare analyses between fecundity and reproductive allocation with either (1) equal sample sizes and a lack of egg size data for K. intermedius or (2) unequal samples sizes. There was no phylogenetic directionality for soldier-to-foundress relative fecundity, egg size or body size–adjusted egg size, or gall size. However, phylogenetic directionality was significant for reproductive allocation, both with the K. intermedius fecundity data (maximum likelihood, p = .042) and without it (p = .041) (Table 3).
|
The same reproductive parameters were tested for covariance with gall size, after accounting for phylogenetic dependence in the reproductive allocation data. Relative egg size did not covary with gall size (maximum likelihood p = .108), whereas fecundity covaried marginally nonsignificantly with gall size (maximum likelihood p = .077) and reproductive allocation covaried significantly with gall size, both with (p = .029) and without (p = .022) the inclusion of the K. intermedius fecundity data.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This study reveals patterns of egg size and fecundity within and between species that help to shed light on intraspecific and interspecific reproductive dynamics in eusocial thrips. The data highlight the importance of two parameters in studies on reproductive allocation: (1) absolute egg size, which has potential implications for competition, maturation, fitness, and survival of offspring and (2) body size–adjusted egg size, which is an indicator of a female's investment in producing an offspring (Smith and Fretwell, 1974
). In particular, egg sizes in K. morrisi stress that fecundity alone can be an insufficient measure of reproductive division of labor and, indeed, makes the difference between interpreting this species as presocial and eusocial (Wilson, 1971; Kranz et al., 200b). Also, egg sizes in K. waterhousei indicate a greater reproductive division of labor than estimated by fecundity alone, and species contrasts reveal a phylogenic rooting and directionality of low to high reproductive allocation "skew" from the basal to derived lineages in the eusocial thrips clade that is not as evident using fecundity data alone (Table 3). Body size–adjusted egg volume and reproductive allocation data were found to covary with gall size more significantly than with fecundity data alone, providing the strongest empirical evidence to date that gall size may be a determinant of reproductive division of labor in eusocial thrips.
The hypothesis that disperser-destined eggs produced by soldiers will be larger than those produced by the foundress, both absolutely and after adjusting for body size, was only partly supported. In terms of absolute egg size, soldiers produced significantly larger disperser-destined eggs than the foundress in K. hamiltoni, K. waterhousei, and K. habrus, whereas egg volume did not differ between castes in K. morrisi (Table 2; Figure 2). In contrast, there was no significant difference in body size–adjusted egg volume between the foundress and soldiers of K. hamiltoni and K. habrus, whereas soldiers had significantly smaller eggs than the foundress in K. waterhousei ( = 0.1: p = .07) and K. morrisi (Figure 3). Combining body size–adjusted egg volume and relative fecundity data of the foundress and soldiers (Kranz et al., 1999, 2001a,b), the average reproductive allocation of a soldier relative to the foundress does not change from that of fecundity in K. hamiltoni and K. habrus but is reduced from about 47% to 35% for K. waterhousei and 110% to 83% for K. morrisi.
Reproductive allocation in eusocial thrips
The relatively large soldier-produced eggs of K. hamiltoni and K. habrus are apparently a function of body size rather than a trade-off with fecundity, as there was no difference in body size–adjusted egg volumes between the foundress and soldiers of these species (Figure 3), although it is possible that the foundress and soldiers differ in their metabolic or nutritional demands during reproduction. Even so, the disperser-destined eggs produced by the soldiers are, on average, larger than those produced by the foundress, and such size variation could potentially lead to developmental, competitive, and fitness differences between soldier and foundress offspring. For K. waterhousei, the larger soldier eggs cannot be accounted for by body size alone if p = .07 is considered significant, which is not unreasonable considering the relatively small sample sizes (Table 1), suggesting a possible trade-off between fecundity and egg size. However, no conclusions can be made from this study to link egg size with fitness optimization in any of the species.
There are two feasible, but not necessarily mutually exclusive, explanations for the possible fecundity and egg size trade-off in K. waterhousei. First, foundress manipulation through protogyny (producing female soldiers first; Kranz et al., 2001a), in conjunction with the energy involved in soldier morphology and defensive behavior, could constrain female soldiers to delay or limit sexual reproduction because they cannot mate until well after eclosion, and in response to this constraint the soldiers could invest more than the foundress in producing a disperser-destined egg. However, it is difficult to understand a motivation for the foundress to constrain soldier reproduction beyond that related to defensive morphology and behavior. The combination of haplodiploidy and sibmating between soldiers means that the foundress will be more related to the female offspring of the soldiers (her granddaughters) than to her own daughters and to her sons more than to her grandsons. Conversely, female soldiers will be more related to their sisters than to their daughters and to their sons more than to their brothers and nephews (Kranz et al., 1999, 2001a). This means that there should be a trade-off for the foundress between ensuring that soldiers defend the gall occupants and facilitating soldiers to produce female offspring, resulting in a possible "reverse conflict" (cf. Trivers and Hare, 1976) over female disperser production, where each caste prefers the other to produce the females (Kranz et al., 2001b).
The second, nonexclusive, explanation is that density-dependent competition, arising from increasing brood size and diminishing gall quality with age, selects for reduced fecundity and, therefore, large eggs. This explanation is reinforced by the phylogenetic contrast studies of Wills et al. (2001) and this paper, which show that gall size correlates negatively with soldier-to-foundress fecundity, and the additional finding in this study that gall size correlates even more significantly with reproductive allocation than it does with fecundity alone (Table 3). The problem with this explanation is that we might then predict a positive relationship between total brood size (and coincident gall age) and body size–adjusted egg size, but this was not found in foundress-only or soldier-only galls. Even so, it is possible that there is an interaction between fecundity constraints, such as those due to soldier defense and the density-dependent competition constraints associated with brood size and gall size and age.
Egg size in galls with the foundress and female soldiers
Mean egg sizes in galls with both the foundress and female soldiers appear to mirror the sex-specific order of soldier eclosion and caste-specific fecundity within each species (Figures 1 and 2). In K. hamiltoni, soldiers are protandrous, so that female soldiers mate and commence reproduction soon after eclosion, and as a group they produce about 85% of the dispersers (Kranz et al., 1999, 2001b). These features are consistent with the mean egg size in galls with both adult types being closest to that of the soldiers. It is unclear why the eggs in soldier-only galls were, on average, smaller than those in galls with both adult types, particularly as there was no relationship between egg volume and total brood size in foundress-only galls. It is possible that the foundress uses soldier reproduction or gall age as an imminent mortality cue and lays large eggs just before she dies as a terminal investment strategy (Clutton-Brock, 1984), or it could be due to the smaller sample sizes in the soldier and foundress galls than in the other gall types (Table 1). More data and investigation are needed to address this.
Soldiers of K. waterhousei and K. habrus are protogynous, such that female soldiers cannot mate until their brothers have eclosed. As a group, soldiers produce about 80% of dispersers in K. waterhousei but about 45% in K. habrus (Kranz et al., 2001a). In K. waterhousei, mean egg size in galls with both adult types were intermediate to those produced in the foundress-only and soldier-only galls, which is consistent with soldier reproduction commencing later than in K. hamiltoni soldiers. Terminal optimal investment by the foundress might also be a factor. In K. habrus, mean egg size in galls with both adult types did not differ significantly to those in foundress-only galls, which is consistent with protogyny and low disperser production by the soldiers. The data imply a lack of terminal investment by the foundress in this species.
Sociality in K. morrisi: a special case
The parity in average per capita fecundity between the foundress and soldiers of K. morrisi was the basis for the argument of Kranz et al. (2001b) that the species is not eusocial by any definition. This parity has been explained in terms of its unusually large galls, which are on average about 3–20 times larger than those of the eusocial species in the clade, and the consequent greater food and space for soldiers to reproduce (Chapman et al., 2002; Kranz et al., 2001b; Wills et al., 2001; Figure 1). What is revealed here, however, is that soldiers of K. morrisi do invest less in reproduction than the foundress. Indeed, while classical definitions of eusociality infer that reproductive division of labor relates to fecundity or brood size (Michener, 1969; Wilson, 1971), the division can and should also apply to total reproductive allocation. In terms of social structure, then, the life history of K. morrisi is consistent with the classical definitions of eusociality (Michener, 1969; Wilson, 1971): offspring of the foundress and soldiers overlap, there is reproductive division of labor, and there is cooperative brood care in the form of defense behavior (Perry et al., 2004). K. morrsi can also be considered eusocial under the narrower definition of Crespi and Yanega (1995) due to the totipotency and irreversibility of the soldier caste.
Evolutionary trends in eusocial thrips
The data from this study reveal that soldier-to-foundress reproductive allocation, but not fecundity or egg size, has significant phylogenetic rooting and directionality from relatively low skew (high soldier reproduction) in the basal lineages to high skew (low soldier reproduction) in the derived lineages (Table 3). While it is clearly the trend that relative fecundity is higher in the basal than in the derived lineages (Figure 1), it is only when body size–adjusted egg size data are incorporated with the fecundity data that the trend is statistically valid. The phylogenetic rooting and directionality for reproductive allocation in eusocial thrips reinforces the suggestions of Kranz et al. (2001b) and Chapman et al. (2002) that eusocial thrips evolved along a subsocial route (Fletcher and Ross, 1985; Wheeler, 1923) and that soldier morphology and behavior evolved before high levels of skew in this clade. Moreover, such directionality suggests a single origin of eusocial thrips, with two losses, rather than two origins and a single loss (Figure 1).
Soldier morphology in thrips evolving under high levels of reproduction is similar to that of aphids (Stern and Foster, 1997) and some termites (Myles, 1986; Thorne, 1997) but contrasts with that of ants, where soldier morphology evolved after that of reproductive division of labor (Bourke and Franks, 1995; Wilson, 1971). The implication is that in thrips, aphids, and termites, the primary selective mechanism for eusociality was that of episodic defense, which still enabled substantial levels of reproduction, in contrast to that of ants and other Hymenoptera, where regular and frequent worker behavior, such as that of foraging and brood care, involve a greater cost for reproduction (Chapman et al., 2002).
Reproductive allocation and fitness
An ongoing challenge in understanding eusocial societies is quantifying and interpreting the relationships between reproductive output, helping behavior, and fitness in subordinate reproductives. Indeed, direct fitness is commonly assumed to be, and measured as, the number of offspring produced (Helms Cahan et al., 2002). In eusocial thrips, it is important to understand whether the larger soldier-produced eggs result in a direct fitness for soldiers that is as high or higher than that for the foundress and what the division in fitness is between the castes. As a first step, comparisons could be made in the sizes of first- and second-instar larvae produced by the foundress and soldiers to determine whether egg size influences posthatching body size. As an indirect approach, correlating within-species foundress body size and brood number could provide insights into body size–associated fecundity. Using an approach similar to that of Charnov (2002) to model reproductive output, offspring body size, gall size and age (and therefore resource quality and availability), and the trade-off between defense and offspring survivorship, rather than collateral survivorship, may be a feasible theoretical approach to understanding direct fitness of the foundress and soldiers in thrips.
The important message from this paper is that fecundity data alone do not necessarily provide a true picture of reproductive investment and division of labor. Studies on egg size in association with fecundity and fitness measures can provide insights into reproductive investment and optimization by the mother and the potential competitive ability of her offspring. Egg size is a parameter that deserves greater attention in studies of reproductive division of labor, particularly in eusocial groups where subordinates have some reproductive ability.
|
ACKNOWLEDGEMENTS |
|---|
Many thanks to Mike Schwarz, Bernie Crespi, and Laurence Mound for encouraging me with this project and to Michelle Guzik for her help with phylogenetic contrasts. Thanks also to Theresa Jones, Mike McLeish, Jim Mitchell, David Morris, and three anonymous referees for their useful comments on the manuscript. The research and manuscript preparation was funded by Australian Research Council grants to M. Schwarz, B. Crespi, and L. Mound (A19702113) and to B.K. and D. Morris (DP0451549).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Andrade MCB, Banta EM, 2002. Value of male remating and functional sterility in redback spiders. Anim Behav 63:857–870.[CrossRef]
Beyer WH, 1987. CRC standard mathematical tables, 28th ed. Boca Raton, Florida: CRC Press.
Bonneaud C, Mazuc J, Chastel O, Westerdahl H, Sorci G, 2004. Terminal investment induced by immune challenge and fitness traits associated with major histocompatibility complex in the house sparrow. Evolution 58:2823–2830.[CrossRef][Web of Science][Medline]
Bourke, AFG, Franks, NR, 1995. Social evolution in ants. Monographs in behavior and ecology. Princeton, New Jersey: Princeton University Press.
Brockelman WY, 1975. Competition, the fitness of offspring and optimal clutch size. Am Nat 109:677–699.[CrossRef]
Brown CA, 2003. Offspring size-number trade-offs in scorpions: an empirical test of the van Noordwijk and de Jong model. Evolution 57:2184–2190.[CrossRef][Web of Science][Medline]
Chapman TW, Kranz BD, Bejah K, Morris D, Schwarz MP, Crespi BJ, 2002. The evolution of reproductive skew in social thrips. Behav Ecol 13:519–525.
Charlesworth B, 1971. Selection in density-regulated populations. Ecology 52:469–474.[CrossRef]
Charnov EL, 2002. Reproductive effort, offspring size and benefit/cost ratios in the classification of life histories. Evol Ecol Res 4:749–758.
Clarke B, 1972. Density-dependent selection. Am Nat 106:1–13.[CrossRef]
Clutton-Brock TH, 1984. Reproductive effort and terminal investment in iteroparous animals. Am Nat 132:212–229.[CrossRef]
Clutton-Brock TH, 1991. The evolution of parental care. Oxford: Oxford University Press.
Crespi BJ, 1992. Eusociality in Australian gall thrips. Nature 359:724–726.[CrossRef]
Crespi BJ, Carmean D, Chapman T, 1997. The ecology and evolution of galling thrips and their allies. Ann Rev Entomol 42:51–71.[CrossRef][Web of Science][Medline]
Crespi BJ, Morris DC, Mound LA, 2004. Evolution of ecological and behavioural diversity: Australian Acacia thrips as model organisms. Canberra: Australian Biological Resources Study and CSIRO Entomology.
Crespi BJ, Mound LA, 1997. Ecology and evolution of social behavior among Australian gall thrips and their allies. In: The evolution of social behavior in insects and arachnids (Choe JC, Crespi BJ, eds). Cambridge: Cambridge University Press; 166–180.
Crespi BJ, Yanega D, 1995. The definition of eusociality. Behav Ecol 6:109–115.
Fletcher DJC, Ross KG, 1985. Regulation of reproduction in eusocial Hymenoptera. Ann Rev of Entomol 30:319–343.
Fox CW, Czesak ME, 2000. Evolutionary ecology of progeny size in arthropods. Ann Rev Entomol 45:341–369.[CrossRef][Web of Science][Medline]
Heins DC, Baker JA, Guill JM, 2004. Seasonal and interannual components of intrapopulation variation in clutch size and egg size of a darter. Ecol Freshw Fish 13:258–265.[CrossRef]
Helms Cahan S, Blumstein DT, Sundström L, Liebig J, Griffin A, 2002. Social trajectories and the evolution of social behavior. Oikos 96:206–216.[CrossRef]
Javois J, Tammaru T, 2004. Reproductive decisions are sensitive to cues of life expectancy: the case of a moth. Anim Behav 68:249–255.[CrossRef]
Kranz BD, Chapman TW, Crespi BJ, Schwarz MP, 2001a. Social biology and sex ratios in the gall-inducing thrips, Oncothrips waterhousei and Oncothrips habrus. Insectes Soc 48:315–323.[CrossRef]
Kranz BD, Schwarz MP, Mound LA, Crespi BJ, 1999. Social biology and sex ratios of the eusocial gall-inducing thrips, Kladothrips hamiltoni. Ecol Entomol 24:432–442.[CrossRef]
Kranz BD, Schwarz MP, Wills TE, Chapman TW, Morris DC, Crespi BJ, 2001b. A fully reproductive fighting morph in a soldier clade of gall-inducing thrips. Behav Ecol Sociobiol 50:151–161.[CrossRef]
Laptikhovskii VV, 2002. Ecology of decapod cephalopod reproduction (Cephalopoda : Teuthida, Sepiida) Zool Zh 81:1319–1328.
Liker A, Reynolds JD, Székely T, 2001. The evolution of egg size in polyandrous shorebirds. Oikos 95:3–14.[CrossRef]
Mappes T, Koskela E, 2004. Genetic basis of the trade-off between offspring number and quality in the bank vole. Evolution 58:645–650.[CrossRef][Web of Science][Medline]
McEdward LR, Morgan KH, 2001. Interspecific relationships between egg size and the level of parental investment per offspring in echinoderms. Biol Bull 200:33–50.
McLain DK, Mallard SD, 1991. Source and adaptive consequences of egg size variation in Nezeara viridula (Hemiptera: Pentatomidae) Psyche 98:135–164.
Michener CD, 1969. Comparative social behavior of the bees. Ann Rev Entomol 14:299–342.[CrossRef][Web of Science]
Morris DC, Schwarz MP, Crespi BJ, Cooper SJB, 2001. Phylogenetics of gall-inducing thrips on Australian Acacia. Biol J Linn Soc 74:73–86.[CrossRef]
Mound LA, 1971. Gall-forming thrips and allied species (Thysanoptera: Phlaeothripinae) from Acacia trees in Australia. Bull Br Mus (Nat His) Entomol 25:389–466.
Mound LA, Crespi BJ, 1995. Biosystematics of two new species of Australian Acacia gall-thrips with soldiers (Insecta; Thysanoptera). J Nat Hist 29:147–157.
Mound LA, Crespi BJ, Kranz B, 1996. Gall-inducing Thysanoptera (Phlaeothripidae) on Acacia phyllodes in Australia: host-plant relations and keys to genera and species. Invertebr Taxon 10:1171–1198.[CrossRef]
Myles TG, 1986. Reproductive soldiers in the Termopsidae (Isoptera). Pan-Pac Entomol 62:293–299.
Pagel M, 1997. Inferring evolutionary processes from phylogenies. Zool Scripta 26:331–348.[CrossRef]
Pagel M, 1999. Inferring the historical patterns of biological evolution. Nature 401:877–884.[CrossRef][Medline]
Parker GA, Begon M, 1986. Optimal egg size and clutch size: effects of environment and maternal phenotype. Am Nat 128:573–592.[CrossRef]
Perry SP, Chapman TW, Schwarz MP, Crespi BJ, 2004. Proclivity and effectiveness in gall defense by soldiers in five species of gall-inducing thrips: benefits of morphological caste dimorphism in two species (Kladothrips intermedius and K. habrus). Behav Ecol Sociobiol 56:602–610.[CrossRef]
Perry SP, McLeish MJ, Schwarz MP, Boyette AH, Zammit J, Chapman TW, 2003. Variation in propensity to defend by reproductive gall morphs in two species of gall-forming thrips. Insectes Soc 50:54–58.[CrossRef]
Roff DA, 2002. Life history evolution. Sunderland: Sinauer Associates.
Santo N, Caprioli M, Orsenigo, S, Ricci C, 2001. Egg size and offspring fitness in a bdelloid rotifer. Hydrobiologia 446:71–74.[CrossRef]
Sargent RC, Taylor PD, Gross MR, 1987. Parental care and the evolution of egg size in fishes. Am Nat 129:32–46.[CrossRef][Web of Science]
Shine R, 1978. Propagule size and parental care: the "safe harbor" hypothesis. J Theoret Biol 75:417–424.[CrossRef][Web of Science][Medline]
Sinervo B, Doughty P, 1996. Interactive effects of offspring size and timing of reproduction on offspring reproduction: experimental, maternal, and quantitative genetic aspects. Evolution 50:1314–1327.[CrossRef]
Smith CC, Fretwell SD, 1974. The optimal balance between size and number of offspring. Am Nat 108:499–506.[CrossRef]
Stern DL, Foster WA, 1997. The evolution of sociality in aphids: a clone's eye view. In: The evolution of social behavior in insects and arachnids (Choe JC, Crespi BJ, eds). Cambridge: Cambridge University Press; 150–165.
Svensson E, Sinervo B, 2000. Experimental excursions on adaptive landscapes: density-dependent selection on egg size. Evolution 54:1396–1403.[CrossRef][Web of Science][Medline]
Thorne BL, 1997. Evolution of eusociality in termites. Annu Rev Ecol Syst 28:27–54.[CrossRef][Web of Science]
Trivers RL, Hare H, 1976. Haplodiploidy and the evolution of the social insects. Science 191:249–263.
Valdimarsson SK, Skulason S, Snorrason SS, 2002. The relationship between egg size and the rate of early development in Arctic charr, Salvelinus alpinus. Environ Biol Fish 65:463–468.[CrossRef]
Vehrencamp SL, 2000. Evolutionary routes to joint-female nesting in birds. Behav Ecol 11:334–344.
Wheeler WM, 1923. Social life among the insects. New York: Harcourt, Brace and Co.
Wills TE, Chapman TW, Kranz BD, Schwarz MP, 2001. The evolution of reproductive division of labor and gall size in Australian thrips with soldiers. Naturwissenschaften 88:526–529.[CrossRef][Web of Science][Medline]
Wills TE, Chapman TW, Mound L, Kranz BD, Schwarz MP, 2004. Description and natural history of Oncothrips kinchega, a new species of gall-inducing thrips with soldiers (Thysanoptera, Phlaeothripidae). Aust J Entomol 43:169–176.[CrossRef]
Wilson EO, 1971. The insect societies. Cambridge, Massachusetts: Harvard University Press.
Winkler DW, Wallin K, 1987. Offspring size and number: a life history model linking effort per offspring and total effort. Am Nat 129:708–720.[CrossRef]
Witting L, 1997. A general theory of evolution by means of selection by density dependent competitive interactions. Aarhus: Peregrine Publisher.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. F Russell and V. Lummaa Maternal effects in cooperative breeders: from hymenopterans to humans Phil Trans R Soc B, April 27, 2009; 364(1520): 1143 - 1167. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||