Behavioral Ecology Advance Access originally published online on June 11, 2004
Behavioral Ecology 2004 15(5):735-741; doi:10.1093/beheco/arh069
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Male morphological variation and the determinants of body size in two Otiteselline fig wasps
Department of Genetics, University of Pretoria, Pretoria 0002, Republic of South Africa
Address correspondence to J. C. Moore, who is now at ICAPB, School of Biological Sciences, University of Edinburgh, Ashworth Laboratories, Kings Buildings, West Mains Rd, Edinburgh EH9 3JT, Scotland, UK. E-mail: jamie.moore{at}ed.ac.uk.
Received 8 July 2003; revised 6 October 2003; accepted 28 October 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The chalcid wasps (Hymenoptera) that develop in fig tree inflorescences (Ficus: Moraceae) have often been used to study alternative reproductive behaviors. However, recent work suggests that such behaviors are more complex than previously thought. We investigated this in Otitesella longicauda and O. rotunda. In addition to known dimorphisms in the two species (each have "religiosa" males that use their mandibles to fight for mates in the fig, and "digitata" males that disperse from the fig to mate), we found that religiosa males below species-specific body size switch points have relatively larger mandibles and are less sclerotized than those above. Thus, they are actually trimorphic. We suggest that the religiosa morph variation is linked to fighter/nonfighter alternative mating behaviors, with small (nonfighting) males having relatively larger mandibles because they also use them to pull females out of their galls before mating. Also, we investigated the determinants of wasp body size, and whether females (foundresses) adjust their offspring allocation strategies according to expected offspring size. We found that wasp size is larger in ovaries near the center of the fig, and more females and fewer religiosa males are laid in such ovaries than in those further away. This probably indicates that foundresses lay females when they are expected to be large because their fitness is more body size–dependent than that of religiosa males. We then discuss the implications of our findings for the study of alternative reproductive behaviors and foundress offspring allocation strategies.
Key words: alternative reproductive behaviors, body size–dependent fitness effects, fig wasps, offspring allocation strategies.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Alternative reproductive behaviors are widespread in nature and are often associated with specialized morphologies (Andersson, 1994
). Theory suggests they may arise as follows: (1) alternative strategies, in which different genotypes with equal average fitness are maintained by frequency-dependent selection; (2) mixed strategies, in which genetically monomorphic individuals adopt different tactics with equal average fitness at probabilities determined by frequency-dependent selection; or (3) conditional strategies, in which genetically monomorphic individuals maximize fitness by adopting different tactics depending on their status relative to others (in this case, tactics have unequal average fitness; Gross, 1996; Maynard Smith, 1982; Parker, 1984). There is still debate over how common each of these are (for differing views, see Brockmann, 2001; Gross, 1996; Shuster and Wade 2003).
One group of species often used to study such behaviors are the chalcid wasps (Hymenoptera) that develop in fig tree inflorescences (Ficus: Moraceae). Wasp ecology is as follows: winged females (foundresses), which as haplodiploids determine offspring sex (Cook, 1993), oviposit internally or externally through the fig wall (the inflorescence, or fig, is a spherical structure with flowers lining the inner wall). Their larvae develop in galled ovaries, in galls induced in the fig wall, or as parasites of other larvae (Compton et al., 1994). Adult males can be (1) winged and nonaggressive, dispersing to find mates; (2) wingless and nonaggressive, mating in the fig or on the surrounding leaves; or (3) wingless and very aggressive, using their mandibles to fight for mates in the fig. Some species have dimorphic nondispersing/dispersing males, which are generally considered to represent alternative strategies evolved to take advantage of mating opportunities offered both inside and outside the fig (Cook et al., 1997; Greeff, 1995; Hamilton, 1979).
Other studies, however, suggest that not all fig wasp alternative reproductive behaviors represent alternative strategies. Pienaar and Greeff (2003a,b) found that Otitesella pseudoserrata morph frequency variation between fig crops is greater than expected under allelic determination, and that patterns in this and two other Otitesella species quantitatively agree with the predictions of a model of optimal offspring (sex and morph) allocation when foundresses oviposit sequentially and know what has previously been laid in the fig. Hence, in these species male morph appears maternally determined, probably because, as larvae do not physically interact, foundresses have more knowledge about mating opportunities in the fig. Also, Bean and Cook (2001) found that Sycoscapter australis (previously considered monomorphic) has two male morphs: above a switch-point, mandible length increases with body size at a higher rate than below. They hypothesized that this is because large males fight for mates (and are selected to maximize mandible length), whereas small males avoid fights and mate with females before their location by fighters and/or while they are being fought over (hence being less selected to maximize mandible length). Because the largest fighting males obtain the majority of copulations in competitive situations (see Greeff and Ferguson, 1999), this implies a conditional strategy in which small males increase their fitness by adopting nonfighting behavior. Whether the same occurs in other species is unknown.
Given the relationship with male morphology, another question arising from these studies is what determines wasp body size. Greeff and Ferguson (1999) suggested that size is positively related to larval development time (which is constrained because figs ripen and are eaten by frugivores after pollinator emergence). If so, we predicted that wasps laid in young figs should be largest. Alternatively, Anstett (2001) found that galled ovary size increases with distance from the fig wall (figs have a multiseriate ovary organization; Verkerke 1989). If wasp size is positively related to gall size (as in cynipid wasps; Stone et al. 2002), we predicted that those laid in ovaries near the center of the fig should be largest. Determinants of wasp size may also be important in terms of foundress offspring allocation. Previous work investigates the effects of clutch size, local mate competition (competition between siblings for mates: Hamilton, 1967), and inbreeding on strategies (for review, see Herre et al., 1997; see also Fellowes et al., 1999; Greeff, 1997; Kinoshita et al., 1998, 2002; Molbo et al., 2003; Moore et al., 2002; Pienaar and Greeff, 2003a,b; West and Herre, 1998), but not whether size effects on offspring fitness are taken into account. In light of their findings on copulation success, Greeff and Ferguson (1999) argued that size has a greater effect on fighting male than female fitness. If so, and assuming they have knowledge of offspring size when ovipositing, we predicted that foundresses should lay fighting males when they are expected to be large. These predictions have yet to be tested.
We investigated these questions in O. longicauda and O. rotunda, two externally ovipositing nonpollinating wasp species that gall flowers in Ficus ingens figs. Previously, these species have been considered to have dimorphic males: fighters (the "religiosa" morph) that mate mostly inside the fig, and dispersers (the "digitata" morph) that mate on the surrounding leaves (see Greeff and Ferguson, 1999). However, Van Noort and Rasplus (1997) suggested that religiosa males exhibit size-dependent morphological variation. We collected morphometric data to investigate whether religiosa males are actually dimorphic. Also, we investigated the determinants of wasp body size, and whether foundresses lay fighting males when their expected size is large.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Are religiosa males dimorphic?
In August 2001, we collected 126 F. ingens figs just before wasp release from three trees at the National Botanical Institute (NBI), Pretoria, Republic of South Africa (RSA) (25°45' S 28°14' E), placed them in mesh-lidded pots, and allowed the wasps to emerge. We then collected the religiosa males and measured their head (as an estimate of body size) and mandible lengths at x20 under a binocular microscope. We measured 342 O. longicauda males, and 184 O. rotunda.
We tested for dimorphism using the method of Eberhard and Guiterrez (1991). First, in each species we examined the relationship between head and mandible length for nonlinearity (a potential indicator of dimorphism). We fitted the partial regression equation (hereafter referred to as model 1):
|
|
i is the regression coefficients, and 
is the random component assuming a normal distribution and homogenous variances. An 2 significantly different from zero (F test) indicates nonlinearity and the possible existence of dimorphism. If nonlinearity was indicated, we then tested the data for the presence of a switch-point, that is, a point at which the allometric relationship between head and mandible length changed. We fitted the partial regression equation (hereafter referred to as model 2):
|
|
is the random component assuming a normal distribution and homogenous variances. We iterated x0 (in 0.01-mm steps) through the range of observed head lengths, defining the switch point as the x0 giving the highest adjusted r2 value. We then examined regression coefficient significance with T tests. Dimorphism is indicated by a significant difference in the slopes either side of the switch point (ß2) or discontinuous variation in mandible length (ß3). Finally, we used an F test to determine whether model 2 explained more variation in the data than a quadratic model of allometry (hereafter referred to as model 3):
|
|
is the random component assuming a normal distribution and homogenous variances. Analyses were carried out in SPSS 11.0. The same methods were used by Bean and Cook (2001).
Determinants of wasp body size and their effect on foundress offspring allocation
The effects of larval development time
We could not simultaneously investigate how larval development time and natal ovary position affect wasp body size and foundress offspring allocation because it was difficult to distinguish figs suitable for natal ovary effects to be quantified (see below). Therefore, to investigate development time effects, in August 2001 we secured nylon bags around 16 groups of five to 10 newly developing figs on a F. ingens tree at the NBI. These bags prevent wasps from ovipositing (Moore, 2001). We then visited the tree until we saw Platyscapa soraria pollinating wasps ovipositing in other figs. When this occurred, we collected figs just before wasp release from another tree, placed them in pots, and allowed the wasps to emerge. We then removed any wasps other than P. sorarie from the pots, after which we took the lids off the pots and placed them in the bags with the experimental figs to ensure the figs were pollinated and began to develop. Five days later, when we first saw Otitisellines ovipositing in other figs on the tree, we removed the bags from four groups of experimental figs. One week later, we replaced the bags and removed them from another four groups. We repeated this procedure, which meant that wasps developing in figs in each set of bags had successively shorter development times, for two more weeks, after which we left the figs until just before wasp release (about 2 months after pollination). We then placed the figs in pots, allowed the wasps to emerge, counted them, and stored the Otitesellines in alcohol. Later, we sorted them by species and to estimate body sizes measured female, religiosa and digitata male head lengths as before. In this experiment and the one below, we treated religiosa males as a single morph. We measured all males collected, but not all females, because some disintegrated during storage.
We analyzed the effects of development time on body size with general linear models (GLMs; in SPSS 11.0). For each offspring type in a species, we fitted a model to the data with the week in which the wasps were laid (i.e., development time) as a fixed factor and natal fig (nested within week) as a random factor. Given possible effects of natal ovary position, in addition for each species we calculated mean offspring type size per fig and fitted models to the data with the week in which the wasps were laid and conspecific number in the fig as covariates. Our reasoning was that in figs with lots of wasps oviposition site competition may lead to ovaries further from the center being used and thus a reduction in mean size (conspecific number was the same for each fig, so we could not carry out a combined analysis on individual size because it would have led to pseudo-replication). We obtained similar results when we replaced conspecific number with the total number of Otitesellines or galling wasps (Otitesellines and P. soraria pollinators) in the fig (potentially, all three species compete for sites).
The effects of natal ovary position
To investigate whether natal ovary position affects wasp body size, in May 2003 we collected figs before wasp release from a F. ingens tree at the NBI. We took the figs to the laboratory, split them open, and located ones containing adult wasps that had not left their ovaries. Female fig flowers consist of an ovary attached to a pedicel that grows out from the fig wall (Verkerke, 1989). There is evidence that pedicels of ovaries containing wasps grow more than those that do not, but pedicel length at wasp maturity is probably still a good indicator of ovary position when foundresses oviposit (Anstett, 2001). We therefore dissected ovaries and their pedicels out of figs, and split the ovaries open. If they contained an Otiteselline, we measured its head length as before and, as an estimate of ovary position, pedicel length (from the base to the point at which the ovary was attached) at x10 under a binocular microscope. We measured wasps from 54 figs in total (not all wasps in a fig were measured because some escaped while others were being collected).
We analyzed the effects of natal ovary position on body size by using GLMs. For each offspring type in a species, we fitted a model to the data with pedicel length as a covariate and natal fig as a random factor.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Are religiosa males dimorphic?
In O. longicauda, religiosa male head lengths ranged from 0.296–0.573 mm, and mandible lengths from 0.067–0.320 mm. Thus, the largest male was about twice the size of the smallest, and had mandibles about four times as long.
2 was significantly different from zero when model 1 was fitted to the data (F1,339 = 291.99, p <.001), indicating a nonlinear allometric relationship between the two traits. The head length (x0) that maximized the adjusted r2 for model 2 was 0.40 mm, corresponding to a mandible length of 0.23 mm (r2 =.80). Mandible length was continuously distributed, but allometric relationships either side of the switch point differed significantly (Table 1). The slope before was steeper than the slope after (Figure 1a). Model 2 explained significantly more variation in the data than did model 3 (F1,338 = 14.99, p <.001), indicating that the morph consists of two types of individual (small males also appear less sclerotized; personal observations). Twenty percent of males were smaller than the switch point.
|
|
In O. rotunda, religiosa male head lengths ranged from 0.332–0.632 mm, and mandible lengths from 0.130–0.367 mm. Thus, the largest male was about twice the size of the smallest, and had mandibles about three times as long.
2 was significantly different from zero when model 1 was fitted to the data (F1,181 = 49.07, p <.001), indicating a nonlinear allometric relationship between the two traits. The head length that maximized the adjusted r2 for model 2 was 0.44 mm, corresponding to a mandible length of 0.29 mm (r2 =.62). Mandible length was continuously distributed, but allometric relationships either side of the switch point differed significantly (Table 2). The slope before was steeper than the slope after (Figure 1b). Model 2 explained significantly more variation in the data than did model 3 (F1,181 = 12.38, p <.001), indicating that the morph consists of two types of individual (again, small males also appear less sclerotized; Moore JC, Pienaar J, and Greeff JM, personal observations). Twenty-four percent of males were smaller than the switch point.
|
Determinants of wasp body size and their effect on foundress offspring allocation
The effects of larval development time
Table 3 shows the head lengths of O. longicauda wasps laid at different times in fig development. Those laid in the first week of the experiment had the longest development times. No digitata males were found in figs from the third week of the experiment. When we analyzed the individual size data with week in which wasps were laid as a fixed factor and fig as a random factor, female (F3,64.79 = 0.19, p = ns; fig, F23,179 = 3.18, p <.001), religiosa male (F3,59.63 = 1.62, p = ns; fig, F21,96 = 0.30, p = ns), and digitata male (F2,8.44 = 2.85, p = ns; fig, F9,7 = 1.73, p = ns) head lengths were independent of week. When we analyzed the mean offspring type size per fig data with week and conspecific number as covariates, female (week, F1,24 = 0.39, p = ns; conspecifics, F1,24 = 0.61, p = ns) and religiosa male (week, F1,22= 2.21, p = ns; conspecifics, F1,22 = 0.47, p = ns) head lengths were independent of both week and conspecific number. Digitata male head lengths were marginally nonsignificantly positively related to both week (F1,9 = 3.75, p <.1, ß = 0.02) and conspecific number (F1,9 = 3.78, p <.1, ß = 0.001).
|
Table 4 shows the head lengths of O. rotunda wasps laid at different times in fig development. When we analyzed the individual size data with the week in which wasps were laid as a fixed factor and fig as a random factor, female (F3,51.11 = 0.45, p = ns; fig, F30,99 = 1.75, p <.05), religiosa male (F3,46.27 = 0.24, p = ns; fig, F28,56 = 1.40, p = ns), and digitata male (F3,4.05 = 0.47, p = ns; fig, F4,1 = 0.96, p = ns) head lengths were independent of week. When we analyzed the mean offspring type size per fig data with week and conspecific number as covariates, female (week, F1,30 = 0.36, p = ns; conspecifics, F1,24 = 0.00, p = ns) and digitata male (week, F1,5= 0.22, p = ns; conspecifics, F1,5 = 0.41, p = ns) head lengths were independent of both week and conspecific number. Religiosa male head lengths were independent of week (F1,29 = 0.86, p = ns) but marginally nonsignificantly negatively related to conspecific number (F1,29 = 3.18, p <.1, ß = –0.002). Thus, in both species wasp size was independent of development time and competitor number in the fig.
|
The effects of natal ovary position
In contrast, we did find relationships between head length and natal ovary position (flower pedicel length). In O. longicauda, female head length was marginally nonsignificantly positively related (F1,4 = 5.24, p <.1; between-fig differences, F17,4 = 6.91, p <.05) and religiosa male head length significantly positively related (F1,32 = 35.48, p <.001; fig, F16,32 = 1.75, p <.05) to pedicel length (Figure 2a,b). We could not examine the relationship in digitata males because only two specimens were collected. In O. rotunda, female (F1,20 = 11.28, p <.01; fig, F17,4 = 4.59, p <.001), religiosa male (F1,10 = 6.16, p <.05; fig, F21,10 = 2.30, p <.01), and digitata male (linear regression F1,5 = 7.89, p <.05: we could not test for between-fig differences because no more than one was found in a fig) head lengths were all significantly positively related to pedicel length (Figure 3a–c). Thus, in both species wasps laid near the center of the fig were larger.
|
|
To investigate whether foundresses took into account the effects of natal ovary position on wasp size when making allocation decisions, we tested whether flower pedicel lengths differed between offspring types. We fitted GLMs to the data with offspring type as a fixed factor and fig as a random factor. In O. longicauda, pedicel lengths differed significantly between offspring types (F2,49 = 24.20, p <.001; fig, F23,49 = 1.51, p = ns) (Figure 4a). Pedicels of flowers in which females were laid were significantly longer than those in which religiosa males were laid (F1,49 = 53.28, p <.001), but did not differ from those in which digitata males were laid (F1,49 = 0.57, p = ns). Pedicels of flowers in which digitata males were laid were marginally nonsignificantly longer than those in which religiosa males were laid (F1,49 = 3.08, p <.1). Significant differences between offspring types were also found in O. rotunda (F2,52 = 6.65, p <.01; fig, F2,49 = 2.21, p <.01) (Figure 4b). Pedicels of flowers in which females were laid were significantly longer than those in which religiosa males were laid (F1,52 = 14.64, p <.001), and marginally nonsignificantly longer than those in which digitata males were laid (F1,52 = 2.81, p <.1). Pedicel lengths of flowers in which religiosa and digitata males were laid did not differ (F1,52 = 0.15, p = ns). Thus, in both species more females and fewer religiosa males were laid in ovaries near the center of the fig than in those further away. Probably because of the small numbers collected, patterns involving digitata males were less clear.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Are religiosa males dimorphic?
We have shown that O. longicauda and O. rotunda religiosa male morphs consist of two types of individual with different allometric relationships between head (estimated body size) and mandible length: the relationship is steeper below species-specific body size switch-points than above (the former are also less sclerotized). As some males develop into dispersers (the digitata morphs), the species therefore actually have trimorphic males (previously, they have been considered dimorphic; see Greeff and Ferguson, 1999
). This is the first time that trimorphism has been described in male fig wasps.
Why are religiosa males dimorphic? Bean and Cook (2001) found body size–dependent male dimorphism in S. australis, and hypothesized that those above the switch-point fight for mates, whereas those below avoid fights and mate females before their location by fighters and/or while they are being fought over. Supporting this in O. longicauda and O. rotunda, the largest males in figs copulate with the majority of females in competitive situations (Greeff and Ferguson, 1999), so small males may increase their fitness by adopting nonfighting behavior. Also, less sclerotization may aid mobility. However, in S. australis small males have relatively smaller mandibles than do large males, probably because selection to maximize mandible length is reduced as a result of not fighting (this is also found in other animals with such behaviors: e.g., beetles: Eberhard and Guiterrez, 1991; mites: Radwan, 1993; earwigs: Tomkins and Simmons, 1996). We found the opposite in our study species. Possibly, this indicates that male morphology is not linked to alternative behaviors: if mandible length determines fight outcome, small males may invest relatively more in their production at a cost in terms of sclerotization. This would suggest they sacrifice the longer lifespan likely to result from increased protection during fights for a greater chance of winning them and obtaining mates. Alternatively (and we believe more likely), there may be other selective pressures on mandible length. Religiosa males pull females out of galls before mating them (Moore JC, Pienaar J, and Greeff JM, personal observations), so nonfighters could still need large mandibles. In contrast, S. australis mate in the fig lumen (Cook J, personal communication) so only fighters need them. Work is now required to investigate whether the two male morphs actually do use alternative behaviors.
If the two male morphs do use alternative behaviors, as with Bean and Cook's (2001) findings in S. australis the body size–dependent nature of the morphological variation implies that they represent alternative tactics within a condition-dependent strategy (see Gross, 1996; Maynard Smith, 1982; Parker, 1984). In these, the tactic used is the one that maximizes fitness given the individual's status (in this case body size) relative to others. Although foundresses may have some knowledge of expected offspring size (see below), the decision about which tactic to use is most likely made by the male before entry into the adult phase: at this point it will be able to estimate its adult size more accurately than its mother can at oviposition. This is contrary to religiosa/digitata male morph determination in O. longicauda and O. rotunda, which seems to be under maternal control because foundresses have more knowledge of mating opportunities in the fig (Pienaar and Greeff 2003a,b). Given other anecdotal accounts of size-dependent variation in male morphology and mating tactics (Idarnes spp.: Bronstein, 1991; Hamilton, 1979; Pereira R. personal communication; Philotrypesis spp: Compton SG, personal communication; Vincent, 1991; Sycorcytes spp: Murray, 1990), conditional strategies are probably common in fig wasps. With the paucity of work in the area (see Brockmann, 2001; Gross, 1996; Shuster and Wade, 2003), these species may therefore be an excellent model system for the development and testing of theory on their evolution and maintenance.
Determinants of wasp body size and their effect on offspring allocation
Contrary to the predictions of Greeff and Ferguson (1999), we found that O. longicauda and O. rotunda body sizes were not related to larval development time. Instead, they were positively related to the distance of the natal flower ovary from the fig wall. This could be because size is less space-constrained in ovaries near the center of the fig (Anstett, 2001), and/or because they contain more resources for larval development.
Also, more females and fewer religiosa males were laid in ovaries near the center of the fig than in those further away (we do not discuss digitata male patterns because of the low numbers collected during the experiments). Given natal ovary effects on wasp size, this is contrary to the predictions of Greeff and Ferguson (1999), who argued that religiosa males should be laid when they are expected to be large because their fitness is more size-dependent than that of females. Possibly, this indicates that foundresses do not adjust their strategy according to expected offspring size. Pienaar and Greeff (2003b) found a good fit between data from the two species and the predictions of an optimal offspring allocation model in which foundresses oviposit sequentially and know only what has previously been laid in the fig. In the model, optimal offspring type depends on the likelihood of successfully mating: first and second arriving foundresses lay females or digitata males because they can disperse if there is no subsequent oviposition, but those arriving later lay females or religiosa males (the latter so they can mate already laid females). Such a strategy could lead to the observed pattern if, owing to either fig growth during development or competition for oviposition sites, only early arriving foundresses can oviposit in the inner ovary layers.
Opposing this, however, fig growth effects would have led to a negative correlation between wasp size and fig age at oviposition even without larval development time effects, and oviposition site competition effects seem unlikely because wasp size was not negatively related to competitor number in the fig. Hence, our findings probably instead indicate that foundresses do adjust their strategy according to expected offspring size, but female fitness is more size-dependent than that of religiosa males. Supporting this, the hypothesized fighter/nonfighter alternative mating tactics (see previously) are likely to reduce size effects on religiosa male fitness. Also, in S. australis the average size of females arriving at trees to oviposit is larger than that of those emerging from their natal figs, suggesting that large individuals survive better to oviposition, and large females have greater longevity (Cook JM, unpublished data). As noted earlier (see Introduction), knowledge of offspring size has been ignored in previous work on foundress offspring allocation. In most cases, understanding its effect on strategies will require new theory to be developed (but for the combined effects of local mate competition and intersexual variation in size/fitness relationships on the optimal sex ratio, see Werren, 1984; Werren and Simbolotti, 1984). In addition, assuming relationships between natal ovary position and wasp size are similar, the way in which foundresses adjust their strategy may differ between species. Contrary to our findings, Murray (1990) found that more males and fewer females of the nonpollinators Philotrypesis pilosa (which has fighting males) and Apocrypta bakeri and the pollinator Ceratosolen marchali solmsi (which both have nonfighting males) were laid in ovaries near the center of Ficus hispida figs than further away. Thus, such theory must be accompanied by investigation of body size/fitness relationships in different species.
In conclusion, our findings suggest that fig wasp reproductive strategies are much more complex than previously thought. The diversity of male alternative mating behaviors within species appears to have been under-estimated, and there are probably several pathways by which the different behaviors have evolved. Also, foundresses seem to have considerable knowledge of future offspring fitness when ovipositing and adjust their sex and morph allocation strategies accordingly. Given the importance of fig wasps as organisms for the study of such questions, further research is thus required.
|
ACKNOWLEDGEMENTS |
|---|
We thank the National Botanical Institute, Pretoria, for permission to work in their gardens; James Cook for access to his unpublished data; and Stuart West and David Shuker for comments on the manuscript. This material is based upon work supported by the National Research Foundation under Grant number 2053809. J.C.M. was also supported by a University of Pretoria postdoctoral fellowship, and J.P. by a NRF studentship.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Andersson M, 1994. Sexual selection. Princeton, New Jersey: Princeton University Press.
Anstett M-C, 2001. Unbeatable strategy, constraint and evolution, or how to resolve evolutionary conflicts: the case of the fig/fig wasp mutualism. Oikos 95:476-484.[CrossRef]
Bean D, Cook JM, 2001. Male mating tactics and lethal combat in the nonpollinating fig wasp Sycoscapter australis. Anim Behav 62:535-542.[CrossRef]
Brockmann HJ, 2001. The evolution of alternative strategies and tactics. Adv Stud Behav 30:1-51.
Bronstein JL, 1991. The non-pollinating wasp fauna of Ficus pertusa: exploitation of a mutualism? Oikos 61:175-186.[CrossRef]
Compton SG, Rasplus J-Y, Ware AB, 1994. African parasitoid fig wasp communities. In: Parasitoid community ecology (Hawkins BA, Sheehan W, eds). Oxford: Oxford University Press; 343–370.
Cook JM, 1993. Sex determination in the Hymenoptera: a review of models and evidence. Heredity 71:421-435.[ISI]
Cook JM, Compton SG, Herre EA, West SA, 1997. Alternative mating tactics and extreme male dimorphism in fig wasps. Proc R Soc Lond B 264:747-754.[CrossRef]
Eberhard WG, Guiterrez EE, 1991. Male dimorphisms in beetles and earwigs and the question of developmental constraints. Evolution 45:18-28.
Fellowes MDE, Compton SG, Cook JM, 1999. Sex allocation and local mate competition in Old World non-pollinating fig wasps. Behav Ecol Sociobiol 46:95-102.
Greeff JM, 1995. Offspring allocation in structured populations with dimorphic males. Evol Ecol 9:550-558.[CrossRef]
Greeff JM, 1997. Offspring allocation in externally ovipositing fig wasps with varying clutch size and sex ratio. Behav Ecol 8:500-505.
Greeff JM, Ferguson WH, 1999. Mating ecology of the nonpollinating fig wasps of Ficus ingens. Anim Behav 57:215-222.[CrossRef][ISI][Medline]
Gross MR, 1996. Alternative reproductive tactics and strategies: diversity within the sexes. Trends Ecol Evol 11:92-98.[CrossRef]
Hamilton WD, 1967. Extraordinary sex ratios. Science 156:477-488.
Hamilton WD, 1979. Wingless and fighting males in fig wasps and other insects. In: Reproductive competition and sexual selection in social insects (Blum MS, Blum NA, eds). New York: Academic Press; 167–220.
Herre EA, West SA, Cook JM, Compton SG, Kjellberg F, 1997. Fig wasp mating systems: pollinators and parasites, sex ratio adjustment and male polymorphism, population structure and its consequences. In: Social competition and cooperation in insects and arachnids: the evolution of mating systems, vol. 1 (Choe J, Crespi B, eds). Cambridge: Cambridge University Press; 226–239.
Kinoshita M, Kasuya E, Yahara T, 1998. More highly female biased sex ratio in the fig wasp, Blastophaga nipponica Grandi (Agaonidae). Res Pop Ecol 40:239-242.
Kinoshita M, Kasuya E, Yahara T, 2002. Effects of time-dependent competition for oviposition sites on clutch sizes and offspring sex ratios in a fig wasp. Oikos 96:31-35.[CrossRef][ISI]
Maynard Smith J, 1982. Evolution and the theory of games. Cambridge: Cambridge University Press.
Molbo D, Machado CA, Sevenster JG, Keller L, Herre EA, 2003. Cryptic species of fig-pollinating wasps: Implications for the evolution of the fig-wasp mutualism, sex allocation and precision of adaptation. Proc Natl Acad Sci USA 100:5867-5872.
Moore JC, 2001. The ecology and evolution of gynodioecious figs and their pollinators (PhD dissertation). Leeds: School of Biology, University of Leeds.
Moore JC, Compton SG, Hatcher MJ, Dunn AM, 2002. Quantitative tests of sex ratio models in a pollinating fig wasp. Anim Behav 64:23-32.[CrossRef]
Murray MG, 1990. Comparative morphology and mate competition of flightless male fig wasps. Anim Behav 39:434-443.[CrossRef]
Parker GA, 1984. Evolutionary stable strategies. In: Behavioural ecology: an evolutionary approach (Krebs JR, Davies NB, eds). Oxford: Blackwell Scientific Publications; 30–61.
Pienaar J, Greeff JM, 2003a. Different male morphs of Otitesella pseudoserrata fig wasps have equal fitness but are not determined by different alleles. Ecol Lett 6:286-289.[CrossRef]
Pienaar J, Greeff JM, 2003b. Maternal control of offspring sex and male morphology in the Otitesella fig wasps. J Evol Biol 16:244-253.[CrossRef][ISI][Medline]
Radwan J, 1993. The adaptive significance of male polymorphism in the acarid mite Calyogliphus berlesi. Behav Ecol Sociobiol 33:201-208.[CrossRef]
Shuster SM, Wade MJ, 2003. Mating systems and strategies. Princeton, New Jersey: Princeton University Press.
Stone GN, Schonrogge K, Atkinson RJ, Bellido D, Pujade-Villar J, 2002. The population biology of oak gall wasps (Hymenoptera: Cynipidae). Ann Rev Entomol 47:633-668.[CrossRef][ISI][Medline]
Tomkins JL, Simmons LW, 1996. Dimorphisms and fluctuating asymmetry in the forceps of male earwigs. J Evol Biol 9:753-770.[CrossRef]
Van Noort S, Rasplus J-Y, 1997. Revision of the otiteselline fig wasps (Hymenoptera: Chalcidoidea: Agaonidae), 1: the Otitesella digitata species-group of the Afrotropical region, with a key to Afrotropical species of Otitesella Westwood. Afr Ent 5:125-147.
Verkerke W, 1989. Structure and function of the fig. Experientia 45:612-622.[CrossRef]
Vincent SL, 1991. Polymorphism and Fighting in Male Fig Wasps (PhD dissertation). Grahamstown: Rhodes University.
Werren JH, 1984. A model for sex ratio selection in parasitic wasps: local mate competition and host quality effects. Neth J Zool 34:81-96.
Werren JH, Simbolotti G, 1989. Combined effects of host quality and local mate competition on sex allocation in Lariophagus distinguendus. Evol Ecol 3:203-213.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||