The obligate relationship between figs and their pollinating fig wasps presents a number of interesting evolutionary questions. Among these questions are those that address the origin and maintenance of a mutualism, as well as the forces of evolution acting within that mutualism. This paper attempts to investigate potential answers to these questions, and argues that the relationship between figs and fig wasps could be viewed as antagonistic coevolution taking place within a cooperative mutualism.
A mutualism can be defined generally as an event where two individuals interact, and both benefit from this interaction. Yu (2001) suggested that the long term maintenance of a mutualism has two requirements. First, there must be reciprocal benefit, and these benefits should be received by the intended parties. Second, it is required that the mutualism is stable across generations, so the interaction must be “re-assembled” repeatedly. The mutualisms between many individuals meet the first condition, but the second is only met in a limited set of interactions (Yu, 2001). The Fig Wasps and their figs are one such case.
The tight obligate relationship between figs and their pollinating wasps has been the subject of much research. With a few exceptions, every species of fig interacts with a species of pollinating wasp. A fig fruit, or syconium, has a number of flowers within it, oriented toward the center of the fruit. The female fig wasps enter a receptive fig through the ostiole, losing their wings in the process, and oviposit in the ovaries of female flowers. In the process, they deposit pollen gathered from their tree of origin and fertilize the female flowers. Since each wasp egg develops at the expense of one fig seed, the fig tree is sacrificing some reproductive potential in order to be pollinated (West and Herre, 1998).
Fig wasps are haplodiploid, and there is maternal control of sex. Fig wasps tend to produce only enough males to fertilize the females (this will be addressed in greater detail below), as the males do not disperse outside their natal fig. Males have substantial mouthparts, and emerge from the flowers first. They then chew a hole in the sides of the galls where the females are developing, mate with them through these holes, and subsequently chew holes through the side of the fig through which females can exit and disperse to other trees. In the process of exiting, females gather pollen either actively or passively (depending on the species), which will be used to pollinate the fig in which they lay their eggs (West and Herre, 1998).
As a result of this system, the female fig wasps that are produced from one fig can be considered to be the male reproductive success of that fig, while the seeds that develop can be considered as the female reproductive success of the fig. Thus, the wasps are necessary for the fig to reproduce, just as the fig is necessary for the wasps to reproduce. This obligate relationship is likely what drives the continued re-assembly of the relationship, and may explain the persistence and prevalence of the interaction.
The origin of such a system is a much more complicated question. It has been suggested that similar mutualisms arose through an antagonistic process in which a non-pollinating parasite lays eggs in a fruit, and the plant selectively aborts that fruit, and so the parasite begins to pollinate the fruit in order to prevent abortion (Jousselin et al., 2003). While this is likely in systems such as the Yucca/Yucca Moth mutualism, it is less apparent in the fig wasp system, as figs are not selectively aborted in this manner (Jousselin et al., 2003).
Westerbergh and Westerbergh (2001) presented a model that stated that any action on the part of the seed predator that caused a net increase in seed production by the plant would result in an initial increase in fitness for that plant, and may be the first step toward mutualism. It has recently been noted that some historically parasitic species of fig wasp (those that do not pollinate) have begun pollinating, and indeed do so as effectively as the original mutualists. They also appear not to over-exploit the resources of the fig, and function almost identically to the original pollinator (Jousselin et al., 2001). This suggests that pollination may confer some benefit to the wasp sufficient to favor its evolution.
After such a first step is taken, a series of subsequent steps must occur in order to cement the relationship between the species. It is often difficult to guess how exactly this process occurred in the past, as we can only observe the result. However, by looking at the traits that maintain the mutualism, we can hypothesize what factors would have favored the evolution of such a trait.
The fig wasp/fig mutualism is characterized by both cooperation and conflict. The necessity of one to the reproduction of the other supports the co-evolution of cooperation to the extent that each species needs to be present at the proper time for the interaction to take place. However, the finer points of the interaction are characterized by conflict, with each evolving to take maximum advantage of the resources the other has to offer. So a case can be made for antagonistic co-evolution within the construct of a cooperative mutualism.
The foundation for the fig wasp/ fig mutualism is necessarily characterized by cooperation. The reproductive cycles of the two are linked, and thus the cycles must be timed accordingly. Most fig trees have synchronous intra-tree flowering, while the inter-tree flowering is highly asynchronous, which ensures that there will be figs flowering throughout the year in order to sustain the population of pollinators (Anstett et al., 1997). There is some evidence to suggest that this synchrony and asynchrony predates the origin of the mutualism (Anstett et al., 1997), and though it may have been important as a first step, it did not evolve as a result of the wasp. However, models suggest that the receptivity period during which a fig can be pollinated has evolved to be longer, so that an enormous population of wasps is not necessary to sustain the figs (Anstett et al., 1997). This receptivity period allows for figs to continue to be pollinated even when they are distant from one another, and may also have important implications for gene flow between fig populations (Anstett et al., 1997).
Fig trees are not distributed in a spatially uniform way, so traits that allow the fig to indicate its location to the wasp and the wasp to detect this signal must also have evolved, and indeed they have. The fig tree produces chemical volatiles that both attract the female wasps to the tree, and stimulate the stereotyped behavior of entering a fig. The high dispersal ability of the wasp (up to 10km) may also have evolved in response to maintenance of low density populations of fig (Anstett et al., 1997).
The active pollination of figs by wasps also has some cooperative elements. Some wasps, rather than gathering and spreading pollen incidentally, actively collect pollen from their natal fig in specialized structures called “pollen pockets” and later actively spread the pollen to fertilize their reproductive fig. This type of system has evolved in other insect/plant mutualisms as a solution to preferential abortion of unpollinated fruit, but this is not the case in figs. Rather, it has been suggested that fertilization stimulates flower growth, and to the extent that this supports wasp development, it is beneficial for the wasp (Jousselin et al., 2003). The benefits of active pollination for the fig tree are obvious. However, the development of the stigmatic platform in figs raises some question about the benefits of active pollination. It is possible that this platform, which facilitates the spread of pollen throughout the fig after the wasps place it, was an adaptation to more efficiently utilize the resource. It is also possible that the platform is a defense against preferential pollination of particular flowers, presumably the flowers into which the wasp plans to oviposit (Jousselin et al, 2003). This second hypothesis would indicate some amount of antagonism, even within the cooperative behavior of active pollination.
Figs and wasps theoretically are in agreement regarding the proportion of females in the wasp’s offspring. Both figs and wasps are benefited by highly female biased broods. Figs prefer females because they are the only sex that disperses and spreads pollen, and wasps prefer females because the more females they produce, the more offspring they have. It is also worth noting that the frequent occurrence of inbreeding often makes female wasps more closely related to their daughters than to their sons as well, again favoring a female biased sex ratio (Moore et al., 2002). Thus, the female wasp should only produce enough sons to fertilize her female offspring and release them from the fig. Though females can produce male offspring without mating, they cannot escape their natal fig without male assistance, and therefore the cost of virginity is much higher than it might otherwise be for both wasp and fig (West et al., 1997). An evolutionary premium is placed on the wasp to optimize the number of females produced while providing sufficient numbers of males.
Despite this cooperation, the mutualism between fig and wasp is also characterized by antagonism stemming from a central disagreement regarding the utilization of the resources, since an increase in wasp eggs corresponds to a decrease in seed production (Herre, 1989)? This conflict has the potential to be the driving force behind continual antagonistic co-evolution, because of its importance to both parties. The wasp would like to lay as many eggs as possible, while the fig would like to retain as many ovaries as possible for seed development (Herre, 1989). When examining this system, it is possible to imagine that the co-evolution between wasp and fig is driven by the wasp finding ways to exploit the fig, and the fig responding in a way to limit the wasp.
One possible defense of the fig against wasp exploitation is the staggering of the ovaries. The fig has ovaries in three locations: near the inner part of the fig, near the outer part, and in between. Fig wasps prefer the ovaries most toward the center of the fig, and preferentially do not oviposit in the outer ovaries (Anstett, 2001). Initially it was hypothesized that fig wasps lacked the capability to oviposit down the longer style that is present on these ovaries (Anstett, 2001), but more recently that hypothesis has been refuted. Instead, it seems that the wasps do not develop as well in these outer ovaries due to physical space constraints. Male wasps also have some difficulty reaching this area to find and mate with the females (Anstett, 2001). Thus, it could be argued that the staggering of ovaries is a defense by the fig against the wasp filling all ovaries with eggs, as it is not reasonable for the wasps to “waste” their eggs on these outer ovaries.
Another possible defense developed by the fig against the wasp is dioecy. Many fig species are monoecious, meaning they have male and female flowers within the same fig. In these figs, the wasp life cycle proceeds in the fashion outlined above. In contrast, a dioecious fig has hermaphrodite trees that function as male, producing pollen and wasps, and female trees that produce seeds, but no wasps. The wasps still enter the female figs and pollinate, but they are unable to lay eggs due to the flower shape (Patel and Hossaert-McKey, 2000). Therefore, species in which this system has evolved have avoided entirely the seed/wasp conflict, as they have retained the pollinating services of the wasp without losing any female seeds. They have restricted the wasp reproduction to male figs, which presumably sustain the wasp population sufficiently in spite of the losses of wasps to female figs (Patel and Hossaert-McKey, 2000).
Since it is known that monoecy is ancestral in figs, and dioecy has evolved separately many times (Anstett et al., 1997), it seems likely that dioecy is an evolutionary solution to overcome the fig’s disadvantage in the mutualism, and actually does result in the fig gaining some advantage. The wasp must be under strong selective pressures to avoid female figs, but it has been demonstrated that male fig maturity is synchronized with female fig receptivity (Patel and Hossaert-McKey, 2000). Males release the wasps when female trees represent a substantial portion of receptive figs, and the wasps have little choice but to pollinate the female figs. It has also been suggested that male and female figs release volatiles that are too similar for the wasp to differentiate, or that the selection on a wasp to find a fig and enter is stronger than the selection to discriminate between figs (Anstett et al., 1997).
It is necessary to revisit the role of sex ratio when discussing conflicts between fig and wasp. Observations indicated to researchers that as foundress (wasps entering one fig) number increased, the sex ratio became less female biased and approached one half. It was long assumed that this was due to the effects of local mate competition, or LMC (West and Herre, 1998). When there is only one foundress, LMC is low- her sons will mate with her daughters for certain, and so there is no advantage to the production of additional sons. As the numbers of foundresses increase, so does the mate competition, thereby increasing the advantage of producing males (West and Herre, 1998). Models assuming both inbreeding and LMC generate reliable predictions that seem to coordinate with the observations of sex ratio, though they tend to conservatively estimate the extent of female bias (Moore et al., 2002). Kinoshita et al. (2002) suggested that this higher degree of observed female bias could be due to the fact that all females do not produce equal clutch sizes, as is often assumed in models, and when clutch sizes vary female bias is increased.
Kinoshita et al. (2002) also investigated the mechanism of decreased female bias as foundress number increases by irradiating female wasps such that their eggs did not develop, though they were oviposited normally. With this procedure, it was possible to ascertain the contributions of each of two foundresses to the total number of wasps in the fig. They found that when the wasps had contact with one another inside the fig, both the order in which they entered and their clutch size significantly affected sex ratio. However, when the wasps do not have contact, only clutch size was significant. They attributed this to two factors: first, the limits in the number of oviposition sites, and second, the order in which females lay eggs of each sex.
Due to the limitation on fig ovaries that are useful to wasps, multiple foundresses must compete for those sites. Because a high premium is placed on eliminating virginity, females lay their male eggs first to ensure males will be present in the resultant brood. After doing this, they are free to lay as many females as they can find sites (Kinoshita et al., 2002). This appears to be adaptive for the wasp, unless oviposition sites are severely limited, which is the case for the second foundress. The second foundress lays her male eggs, and then cannot lay as many females as she otherwise would because the sites are all utilized. The result of this process is what appears to be a heavily male-biased sex ratio, when observing all offspring emerging from the fig without regard to parentage (Kinoshita et al., 2002).
In the context of conflict, this means that the defenses of the fig against wasp exploitation are producing effects that are neither beneficial for the wasp nor the fig. An increasingly male-biased ratio does not benefit the wasp significantly, nor does it benefit the fig. So traits that are advantageous for each are less advantageous when viewed in concert. This example illustrates nicely a potential trade-off of a mutualism: what is optimal to each may not be optimal for the interaction.
It is also important to note that sex ratio has other selection factors acting on it in some species. Male bias can be increased when there are high levels of fatal combat between males within a fig, or when males evolve dispersal ability. Fatal combat would require more male bias to ensure fertilization of females (Reinhold, 2003), while male dispersal would eliminate much of the female advantage (Greeff, 2002). Herre (1993) also modeled the virulence of parasites of fig wasps, and noted that the vertical transmission of parasites in a single foundress system favors decreased virulence, while the horizontal transmission in multiple foundress systems favors increased virulence. In theory, this could place some selective advantage on being a single foundress, strengthening the already present selection to be the first foundress. Clearly these factors must also be considered when attempting to examine the role of the sex ratio in the mutualism.
When viewed holistically, it is clear that the fig/wasp system is an excellent example of a coevolved mutualism, founded in cooperation. However, when attention is focused toward the mechanisms acting within that system, conflict appears to be the driving force. Further studies into the relationship between fig and wasp will likely illuminate exactly what roles conflict and cooperation play in the evolution of these species.
Literature Cited
Anstett, M. C. 2001. Unbeatable strategy, constraint and coevolution, or how to resolve evolutionary conflicts: the case of the fig/wasp mutualism. Oikos. 95: 476-484.
Anstett, M.C., M. Hossaert- McKey and F. Kjellberg. 1997. Figs and fig pollinators: evolutionary conflicts in a coevolved mutualism. TREE. 12(3): 94-99.
Greeff, J. M. 2002. Mating system and sex ratio of a pollinating fig wasp with dispersing males. Proc. R. Soc. Lond. B. 269: 2317–2323.
Herre, E. A. 1989. Coevolution of reproductive characters in 12 species of New World figs and their pollinating wasps. Experientia 45: 637-647.
Herre, E. A. 1993. Population structure and the evolution of virulence in nematode parasites of fig wasps. Science. 259: 1442-1445.
Jousselin, E., J. Y. Raspius and F. Kjellberg. 2001. Shift to mutualism in parasitic lineages of the fig/fig wasp interaction. Oikos. 94(2): 287-294.
Jousselin, E., M. Hossaert-McKey, E. A. Herre and F. Kjellberg. 2003. Oecologia. 134: 381-387.
Kinoshita, M., E. Kasuya and T. Yahara. 2002. Effects of time-dependent competition for oviposition sites on clutch sizes and offspring sex rations in a fig wasp. Oikos. 96: 31-35.
Moore, J. C., S. G. Compton, M. J. Hatcher and A. M. Dunn. 2002. Quantitative tests of sex ratio models in a pollinating fig wasp. Animal Behaviour. 64: 23-32.
Patel, A. and M. Hossaert-McKey. 2000. Components of reproductive success in two dioecious fig species, Ficus exasperata and Ficus hispida. Ecology. 81(10): 2850-2866.
Reinhold, K. 2003. Influence of male relatedness on lethal combat in fig wasps: a theoretical analysis. Proc. R. Soc. Lond. B. 270: 1171–1175.
West, S. A. and E. A. Herre. 1998. Stabilizing selection and variance in fig wasp sex ratios. Evolution. 52(2): 475-485.
West, S. A., E. A. Herre, S. G. Compton, H. C. J. Godfray and J. M. Cook. 1997. A comparative study of virginity in fig wasps. Animal Behaviour. 54: 437-450.
Westerbergh, A. and J. Westerbergh. 2001. Interactions between seed predators/pollinators and their host plants: a first step toward mutualism? Oikos. 95: 324-334.
Yu, D. W. 2001. Parasites of Mutualisms. Biological Journal of the Linnean Society. 72: 529-549.