Conflicts of interest in social insects: male production in two species of Polistes.
Reproduction within a colony is typically monopolized by comparatively few females, called queens (Hamilton 1964; Wilson 1971). However, how these few individuals can keep the workers from reproducing is not well understood, especially when there are large numbers of individuals per colony (West-Eberhard 1969; Wilson 1971). For production of female reproductives in the Hymenoptera, relatedness sometimes provides the answer. Because of the haplodiploid genetic system (where females are diploid and develop from fertilized eggs, while males are haploid and develop from unfertilized eggs), a worker is less related to her own daughters (r = 1/2) than to her full sisters (r = 3/4; Hamilton 1964). But this solution never works for male production; a worker is always more related to her own sons (r = 1, by the regression measure of relatedness) than to any other male relative. Four other possibilities for why workers do not re-produce are (1) the queen physically prevents workers from reproducing; (2) workers are incapable of laying eggs because they are truly sterile; (3) workers voluntarily refrain from laying eggs because the costs of conflict are too high; and (4) workers control each other's reproduction (Queller and Strassmann 1998).
The fourth alternative is based on the theory of worker policing (Starr 1984; Woyciechowski and Lomnicki 1987; Ratnieks 1988), which holds that under some conditions workers protect their own genetic interests by preventing other workers from laying eggs. Just because each worker favors her own sons, it does not mean she will favor the sons of other workers. If a worker is more related to the queen than she is to other workers, then other things being equal, she should try to suppress other workers' male production in favor of the queen workers. Of course, other things may not be equal. In particular, costs of conflicts may alter the outcome of conflicts. But it is nevertheless worth testing the simplest hypothesis first: whether relatedness is the key factor in suppressing worker reproduction.
Worker policing may explain how reproduction can be limited to a single queen even in large colonies where physical domination by one queen is unlikely. Honeybees are an example where the mating behavior of the queen favors worker policing. Queens mate over 10 times, so workers are more closely related to brothers (males produced by the queen) than to the half-nephews produced by other workers (Adams et al. 1977; Page and Metcalf 1982). Ratnieks and Visscher (1989) demonstrated that worker honeybees effectively detect and destroy other workers' eggs, allowing only the queen's eggs to develop.
The two conditions determining whether workers should police (relatedness to queens and relatedness to other workers) coupled with the two possible sources of males (queens or workers) define the four possibilities shown in the cells of Figure 1. In an interspecific test, policing theory is best supported if species cluster in cells A and D, in which workers are more related to the class that produces the male eggs. However, any individual species in either of those cells is also consistent with another mode of control: queen control if queens are laying and individual worker control if workers are laying. However, any species falling in one of the other two cells (B and C) is inconsistent with policing theory, at least the simplest version that posits that relatedness alone determines the outcome because costs and benefits are equal. If workers are more related to queens but workers lay (outcome B), it supports control of male production by individual workers rather than the worker collective. If the queen lays the male eggs when workers are more related to other workers than to the queen (outcome C), then only the queen's interests are served, supporting the hypothesis of queen control.
This logic might also apply at the level of individual colonies. If, workers can assess relatednesses on their own colony, then we might expect them to make different choices. For example, if workers collectively control male production, colonies might be divided between outcomes A and D according to their worker relatednesses. If however, workers have no information about how their colony relatednesses differ from the mean, then selection will act on average relatedness alone.
In this study we investigated the worker-queen conflicts in male egg production in two species of the paper wasp Polistes (Vespidae). Polistes is an interesting study organism for this question for three reasons. First, Polistes lacks caste dimorphism. Workers can mate and become fully functional queens so they are certainly reproductively capable (Strassmann and Meyer 1983). Second, species of Polistes form small enough colonies that the queen may be able to maintain complete reproductive control by herself. Therefore, a finding supporting worker control would suggest that policing applies even in small colony species. Third, there appears to be variation among species in both relatedness structure (Strassmann et al. 1989) and in potential for worker laying of males, though the latter has not been extensively studied. Unmated workers in some species of Polistes have been reported to lay male eggs when the queen is removed or her access to the brood is restricted (Miyano 1980; West-Eberhard 1986; Russina et al. 1993). Worker laying in queenright colonies has been investigated in six species of Polistes (P. jadwigae, P. canadensis, P. metricus, P. dominulus, P. nimphus, and P. chinensis) using observations or allozymes (Metcalf and Whitt 1977; Miyano 1980; Strassmann and Meyer 1983; Russina et al. 1993). Worker laying was observed only in P. chinensis (Miyano 1980; Russina et al. 1993), where it was a rare event. The great variation in colony sizes and life histories means that circumstances favoring worker laying may occur in some and not in other species (Reeve 1991).
Worker policing has not yet been studied in Polistes, but relatednesses have been estimated in a number of species from allozyme data (Strassmann et al. 1989). In that study, relatedness among end-of-season females was low in P. bellicosus (0.338 [+ or -] 0.148) and high in P. dorsalis (0.607 [+ or -] 0.079). We selected these two species for a study of policing in part because their relatedness differences might lead them to differ in worker policing. However, these estimates of relatedness had large standard errors and therefore should be considered tentative. Much more precise estimates will be obtained by the use of DNA microsatellite markers, which are highly polymorphic (Queller et al. 1993).
We used microsatellite genotypes for three purposes. First we used them to estimate the relatednesses that allow us to make a prediction about policing. Specifically, we needed to know if workers are more related to other workers or to queens, because whichever is higher should be preferred by workers as the source of males. (With the regression measure of r, which we use in this paper, worker relatedness to the males equals their relatedness to the males' mothers; with the life-for-life measure relatedness to sons is halved, but the relative values of workers and queens are unchanged; Hamilton 1972; Grafen 1986.) Second, microsatellites are needed to determine the sex of the brood. Males are haploid, so they possess only one allele at every locus. Finally, we used the microsatellite genotypes to test the policing prediction by determining whether workers or queens actually produce the males.
MATERIALS AND METHODS
We marked wasps of P. bellicosus and P. dorsalis in native prairies at Brazos Bend State Park, Texas, for a week in late September 1994. We collected the nests at dawn, when most of the wasps are on the colony and inactive. For a few days after collection, we checked all nest sites for presence of uncollected wasps. We collected 12 colonies of P. bellicosus and six of P. dorsalis. The collected colonies for both species were located in a 4-ha plot, with typically 5-10 m separating neighboring colonies. Adult males had begun to emerge on some colonies, but most were still in the brood stage. We kept collected colonies (adults and brood contained in the nest) [TABULAR DATA FOR TABLE 1 OMITTED] alive on ice for transportation to the lab and then preserved them in a -80 [degrees] C freezer.
We dissected all 170 adult females of P. bellicosus and 87 adult females of P. dorsalis, and evaluated their reproductive status (insemination, ovarian development, and presence of eggs, Table 1). For the queens (females with ovarian development and inseminated), we opened the spermathecae and removed the sperm to genotype it (Strassmann et al. 1996).
DNA Microsatellite Genotyping
We extracted DNA from the thoraxes of adult females and mature pupae and from the entire body of young pupae and eggs. We followed standard DNA extraction protocols as in Strassmann et al. (1996).
For the DNA analysis of P. bellicosus, we used 321 specimens that included 114 adult females and 207 brood. This included all adult females for 10 of the colonies. For the largest colonies (ID numbers 8 and 9) we genotyped the queens and samples of 10 workers. We haphazardly selected up to 10 pupae and 10 eggs from each colony to sample the oldest and youngest brood (Table 1A).
For P. dorsalis, we used all 216 specimens that included 87 adult females, eight adult males and 121 brood. As with P. bellicosus, for the brood we haphazardly selected up to 10 pupae and 10 eggs from each colony (Table 1B). Inclusion of adult males in the male parentage analysis could result in error if males had come from another colony. We included them in the study of P. dorsalis to increase the sample sizes. Removing those males from the analyses did not cause any change in the results.
We used seven pairs of trinucleotide microsatellite primers for P. bellicosus (Pbe80AAC, Pbe203AAG, Pbe269AAG, Pbe411AAT, Pbe424AAT, Pbe440AAT, Pbe492AAT) and eight pairs for P. dorsalis (Pbe80AAC, Pbe102TAG, Pbe203AAG, Pbe205AAG, Pbe269AAG, Pbe411AAT, Pbe424AAT, Pbe492AAT). Primers and amplification conditions are given in Strassmann et al. (1997). These microsatellites had between four and 21 alleles, and heterozygosities from 0.61 to 0.92. We used them to genotype the queen(s), workers, brood, and the queen's sperm from each colony. Sperm genotyping did not succeed for P. dorsalis.
We used a 10-[[micro]liter] PCR reaction, and ran PCR products from two different loci with different allele sizes (which ranged about 100 bp different) together in each lane on denaturing polyacrylamide gels using an m13 sequencing reaction as a size standard. Once autoradiographs were obtained, two people scored them and entered the data. Any discrepancies were rechecked against the autoradiograph and, if necessary, the sample was rerun.
While scoring the genotypes for P. bellicosus and assigning progeny to the correct mother, we found nonamplifying alleles or null alleles (Callen et al. 1993; Pemberton et al. 1995) for two of the loci in two different colonies. Null alleles occur when the primer binding site is so different that the primer does not bind and DNA amplification fails. In colony 6 we suspected a null allele at locus Pbe80AAC for the queen's genotype, which would explain why some of her female offspring appeared homozygous for the paternal allele and why some males had no amplification product. By modifying the PCR reaction, (decreasing the annealing temperature by two degrees) we were able to amplify the null allele. The other null was in colony 7 for the locus Pbe492AAT, where we inferred the null allele for the paternal genotype. All the female offspring appeared homozygous for one or the other of the two alleles possessed by a queen who, at all other loci, was consistent with being the mother. Therefore, we assigned the null allele an additional character distinct from the other known alleles at these loci.
Distinguishing Sex of Brood
We used microsatellite markers to distinguish diploid females from haploid males in all genotyped brood. Old pupae could also be divided into males and females by counting antennal segments (males have 13, whereas females have 12). We considered individuals with only one allele at every locus to be hemizygous males. Because observed heterozygosities were high (average of 81% in P. bellicosus and 75% in P. dorsalis), the probability that a female would be misassigned as a male because she was homozygous at all loci was very small in both species (7.14 x [10.sup.-6] for P. bellicosus and 4.76 X 10 6 for P. dorsalis, assuming the loci are independent).
We used the microsatellites to estimate relatednesses with the program Relatedness, version 4.2b (by K. F. Goodnight and D.C. Queller). We report the regression measures that this program estimates. We weighted colonies equally in the relatedness estimations and estimated standard errors by jackknifing over colonies, except for single-colony estimates, which were jackknifed over loci. We calculated 95% CI assuming that the pseudovalues followed a t-distribution (Queller and Goodnight 1989).
Maternity Assignments and Male Production
Maternity assignments are sometimes carried out by excluding all possible nonmothers. This approach was not adopted for determining parentage of males in this study because exclusion of a large number of close relatives is difficult, and because some of the individuals who might have been mothers (particularly workers) died before collection or were not genotyped. Instead, we adopted a maximum-likelihood approach, L, finding the most likely value of the parameter O, the fraction of males produced by queens as opposed to workers (Hastings et al., in press). For each of the male-producing colonies, the relative likelihood of values of Q was calculated as:
[Mathematical Expression Omitted] (1)
for all values of Q between zero and one (actually we sample Q values at small intervals). K is a multinomial constant that never has to be calculated because it multiplies all Ls by a constant, leaving unaffected which is the maximum. For each male allele considered in turn, [f.sub.qi] and [f.sub.wi] are the frequencies of that allele in the queen and in the workers. Overall likelihoods for all colonies were calculated as the product of the colony-specific likelihoods for a given value of Q.
Note that the equation pools the frequencies of all the queens and all the workers. By aggregating the worker alleles into a common set of frequencies, this method has the advantage of allowing for plausible worker genotypes that we did not sample. For example, if only two workers are sampled with two-locus genotypes of AA/CD and AB/DD (loci separated by slashes) neither could produce a male with a two-locus genotype of B/C. However, they could easily have a missing sister worker with a genotype that could produce that male (e.g., AB/CD). Our method allows for this possibility by drawing from a worker pool consisting of 3/4 A and 1/4 B for the first locus and 1/4 C and 3/4 D for the second.
For testing policing theory, the queens of interest are those who were producing female progeny at the time when male progeny were being produced (eggs and/or pupae, depending on the colony). These might be a subset of the queens collected (there was more than one inseminated female on some P. bellicosus nests) or they might be queens that were not collected because they had died before the time of collection. The queens that were reproductively active at the appropriate time were identified as follows. First, all nonqueen females (workers and brood) were aggregated into full sister groups that were as large as possible, subject to the requirement that at each locus, all members of the sibship must share one allele (the paternal one), and collectively possess no more than two additional alleles (the maternal ones). We did this by first getting a rough sort using the sorting feature of the program Kinship 1.1 (by K. F. Goodnight and D. C. Queller) and then manually checking the genotypes. In P. bellicosus, this procedure was effective in the sense that whenever a full sister group matched the genotype of a living queen, they also matched her genotyped sperm. Sperm did not amplify well in P. dorsalis, so this double check was not possible.
When a full sister group did not match any living queen, we used their genotypes to determine the genotype of their dead mother. When all daughters in a sibship are heterozygous, it cannot be determined which allele was maternal and which was paternal, so such loci were entered as missing for the mother in the likelihood analysis. For P. bellicosus, this occurred for a mother in colony 1 (one locus missing) and another in colony 8 (two loci missing). For P. dorsalis, colony 1 had to be excluded from the analysis because it had only one female pupa and one female egg, which was not enough to specify the maternal genotype.
When a dead mother is heterozygous and happens to pass on only one of the alleles to all of her daughters, the other allele would be missed. This chance was small because sibships were generally large. If we ever missed such an allele, it would create a bias against assigning males to the queen.
Colony Size Characteristics
In Polistes bellicosus, colonies had from four to 60 adult females. Male brood were found in all but two colonies (colonies 5 and 7). The majority of the males in the brood were at the pupal stage; only five of the eggs were male (Table 1A). In addition, for two of the largest colonies (colonies 8 and 9) [TABULAR DATA FOR TABLE 2 OMITTED] a total of 17 adult males were also collected (Table 1A).
In P. dorsalis, the colonies had from six to 26 adult females, a range lower than that for P. bellicosus. On two of the largest colonies, a total of eight adult males also were collected (Table 1B). In P. dorsalis, males were also present among both the pupae and the eggs (Table 1B).
Polistes bellicosus colonies had 31 inseminated females in 10 of the 12 colonies (Table 1A). Of those 31, only 16 also had ovarian development. The other 15 females may have been older queens whose ovaries had regressed or the next season's queens who had already been inseminated. In many of them the presence of stored fat, necessary for overwintering, suggested that they were the next season's queens. We did not collect any queens on colonies 1 and 6. Four colonies (colonies 7, 8, 9, and 12) had more than one current queen inseminated with developed ovaries (Table 1A). No uninseminated female had mature eggs in her ovaries.
The P. dorsalis colonies had one inseminated queen per nest (Table 1B). Besides these queens, no ovarian development or fertilization was found in any other female. Therefore, all other adult females were classified as workers.
In both species, we sorted offspring of each colony into full sister groups. In general, these groups could not have been produced by the same mother, indicating that most queens are singly mated. One exception was found on colony 9 of P. bellicosus, which had two full sibships, with 15 and three females, respectively, that could have come from the same mother mated to two males. However, P. bellicosus sperm amplifications also showed that queens mated only once.
Predicted Worker Preferences from Relatednesses
Table 2 shows the relatednesses, standard errors jackknifed over loci, and the 95% CI for the two species of Polistes. With these relatednesses we can determine the workers' preferences for who should be laying the male eggs, and provide a preliminary indication of who is actually doing so. If relatedness among workers is lower than worker relatedness to the queen(s), workers should prefer for the queen to lay male eggs; otherwise they should allow other workers to lay. For both species, the relatedness among workers is higher than the relatedness of workers to queens (Table 2). The paired difference test (Queller 1994) showed that there is a significant difference between them (0.21, 95% CI 0.03 [less than] [r.sub.workers] - [r.sub.worker-queen] [less than] 0.38, in P. bellicosus, and 0.20, 95% CI 0.06 [less than] [r.sub.workers] - [r.sub.worker-queen] [less than] 0.33 in P. dorsalis).
Who is actually producing the males? The relatedness of workers to males was 0.41 in P. bellicosus and 0.58 in P. dorsalis, which is nearly identical to the relatedness of workers to queens (0.40 and 0.54 for P. bellicosus and P. dorsalis, respectively). This similarity would be expected if the queens were producing the males. The paired difference between worker-worker relatednesses minus worker-male relatedness was significantly greater than zero in both species: 0.19 (95% CI 0.02 [less than] r - r[prime] [less than] 0.41) in P. bellicosus; and 0.16 (95% CI of 0.02 [less than] r - r[prime] [less than] 0.30) in P. dorsalis. This suggests that workers are not producing the males [ILLUSTRATION FOR FIGURE 2 OMITTED].
For the likelihood analysis, we used only the subset of queen genotypes that were producing female progeny at the time male progeny were produced. Usually this was only one queen per nest. For P. bellicosus, this queen was among those collected for five colonies (colonies 2, 3, 4, 10, and 11), and had previously died in three colonies (colonies 1, 6, and 12). Colony 8 had two laying queens, one collected and one dead, and colony 9 had two laying queens both of which were collected. Figure 3a shows the likelihood curves for the proportion of queen-produced males in P. bellicosus. Clearly, most males are queen produced; that is, male-destined eggs are laid by the same class that is laying female-destined eggs. For nine of the 10 colonies, the maximum likelihood was that all males are queen produced. For all colonies combined (heavy line in [ILLUSTRATION FOR FIGURE 3A OMITTED]), the maximum likelihood is that 99% of males are produced by the queens. One hundred percent queen production has zero likelihood because one male egg in Colony 8 could not have been produced by either of the two active queens. It is uncertain whether that male was produced by a worker or by a queen unrepresented among the female progeny, but we can safely conclude that very few males are worker produced. The combined likelihood drops rapidly from Q = 0.99; any worker laying of 12% (Q = 0.88) or more is 20 times less likely, and for 18% (Q = 0.82) it is more than 500-fold lower.
For P. dorsalis [ILLUSTRATION FOR FIGURE 3B OMITTED], five of the six colonies had one collected queen who was laying female eggs at the time when male brood were laid (Colony 1 had only one female in each brood category and was eliminated from the likelihood analysis). The likelihood curves are shown in Figure 3b. All five of the colonies showed a maximum likelihood of complete queen production of males (thus this result held for colonies with and without adult males in the sample). The combined curve (heavy line) drops very steeply. For example, any worker laying of 3% (Q = 0.97) or more is 20 times less likely, and for 10% (Q = 0.9) it is more than 30,000 times lower.
Reproduction in Polistes may be regulated by queens or by workers, acting collectively or individually. While queens should always prefer to reproduce, whether workers prefer themselves as a class to reproduce or prefer the queen to reproduce, depends on their relatednesses [ILLUSTRATION FOR FIGURE 1 OMITTED]. For both Polistes studied (P. bellicosus, P. dorsalis), worker-worker relatednesses were significantly higher than worker-queen relatednesses ([ILLUSTRATION FOR FIGURE 2 OMITTED], Table 2). Worker policing theory therefore predicts that, other things being equal, workers should prefer workers over the queens as the source of males. However, in both species all or virtually all of the males were produced by the queens. Thus, of the four possibilities in Figure 1, both of these species support queen control (cell C).
Why do workers not lay male eggs even though policing theory predicts that they should? There are several possibilities. First, it may be due to the colony size. In general, the colonies for these two species are small (four to 10 adult females). It is more reasonable to believe that the queen might maintain reproductive control through physical dominance over a small number of other individuals (West-Eberhard 1986; Ratnieks and Reeve 1992). However, two of the colonies of P. bellicosus (colonies 8 and 9) were considerably larger than the others, with about 60 adults per colony. Both nests had two queens contributing to the brood. This may be a consequence of decreased ability of one individual to control the nest, but even in these two cases, the males were produced by queens rather than workers.
A second possibility is that the relatednesses we measured were atypical and that worker-worker relatedness is normally in the worker policing range. An allozyme study of late-season females (Strassmann et al. 1989) yielded a relatedness for P. bellicosus of 0.34 (95% CI 0.04-0.64), and for P. dorsalis of 0.64 (95% CI 0.44-0.77). This suggests that P. bellicosus, but not P. dorsalis, may sometimes fall in the low-relatedness range where worker policing would be favored. However, the low P. bellicosus relatedness for that study had very large standard errors, and a third study of P. bellicosus is consistent with the high relatedness we found in this study. In that study of early-season P. bellicosus colonies, from the same locality as the present study, relatednesses were 0.54 for workers and 0.67 for foundresses (Field et al., in press). This shows that the best evidence we have is that the relatednesses for both species are typically high.
We have assumed that if workers did lay male eggs, they would do so at random with respect to relatedness. If workers were differentially related to different classes of workers (for example old versus young), they could have the option of aiding the class to which they are most related on average. This would raise the average relatedness to worker-produced males, and make our finding of queen-laying even more contrary to worker relatedness interests. The same would be true if workers could actually recognize and aid their full sisters, though this seems unlikely based on current evidence (Keller 1997).
A third possibility is that allowing the queen to lay all male offspring may help guarantee the colony's success by avoiding costs of conflicts (Cole 1986; Ratnieks and Reeve 1992). Such costs could include direct aggression, recognition errors, or decreased foraging as workers compete to lay eggs. When conflicts are costly, selection can favor settling them via some convention, even if that convention is arbitrary (Maynard Smith 1974; Pollock 1996). Thus, conflict costs could result in the convention "only queens lay male eggs," even when workers could obtain higher relatedness to brood if workers laid the male eggs. Of course, this does not explain why the convention is not "only workers lay male eggs," but there are some possible reasons. First, if the queen is historically the source of all eggs, then the easiest route to a convention is to leave things as they are. Second, because the queen is the source of female eggs, she may be able to lay at a lower physiological cost. Third, the worker-laying solution does not completely remove conflict because workers will still compete against each other, whereas investing reproduction in a single queen would remove conflict.
A lack of worker control has been found in another context for Polistes. In P. annularis, when the original queen dies, she is succeeded by a subordinate foundress, if one is present, and then by an older worker (Hughes et al. 1987). Microsatellite analysis of the relationships showed that workers should have preferred the opposite; on average, workers are more related to the progeny of a young worker than to those of an old worker and more related to the progeny of a random worker than to those of a subordinate foundress (Queller et al. 1997).
In conclusion, we have found no evidence for worker control. It is not clear whether the cause lies in effective dominance by the queen or in the adoption of conventional settlements due to high conflict costs. But it is clear that, for whatever reason, workers do not collectively enforce their preferred outcome with respect to relatedness.
We thank K. Goodnight for computer programming, J. Field for fieldwork, and J. Eberle and two anonymous reviewers for comments on the manuscript. Financial support was provided by a research grant (DEB-9510126) and a postdoctoral fellowship to EA (BIR-9406843) from the National Science Foundation.
ADAMS, J., E. D. ROTHMAN, W. E. KERR, AND Z. L. PAULINO. 1977. Estimation of the number of sex alleles and queen matings from diploid male frequencies in a population of Apis mellifera. Genetics 86:583-596.
CALLEN, D. E, A.D. THOMPSON, Y. SHEN, H. A. PHILLIPS, R. I. RICHARDS, J. C. MULLEY, AND G. R. SUTHERLAND. 1993. Incidence and origin of "null" alleles in the (AC)n microsatellite markers. Am. J. Hum. Genet. 52:922-927.
COLE, B. J. 1986. The social behavior of Leptothorax allardycei (Hymenoptera: Formicidae): time budgets and the evolution of worker reproduction. Behav. Ecol. Sociobiol. 18:165-173.
FIELD, J. E, C. R. SOLIS, D. C. QUELLER, AND J. E. STRASSMANN. In press. Social and genetic structure of paper-wasp co-foundress associations: tests of reproductive skew models. Am. Nat.
GRAFEN, A. 1986. Split sex ratios and the evolutionary origins of eusociality. J. Theor. Biol. 122:95-121.
HAMILTON, W. D. 1964. The genetical evolution of social behaviour. 1. J. Theor. Biol. 7:1-16.
-----. 1972. Altruism and related phenomena, mainly in social insects. Annu. Rev. Ecol. Syst. 3:193-232.
HASTINGS, M. D., D. C. QUELLER, F. EISCHEN, AND J. E. STRASSMANN. In press. Kin selection, relatedness, and worker control of reproduction in a large-colony epiponine wasp, Brachygastra mellifica. Behav. Ecol.
HUGHES, C. R., M. O. BECK, AND J. E. STRASSMANN. 1987. Queen succession in the social wasp Polistes annularis. Ethology 76:124-132.
KELLER, L. 1997. Indiscriminate altruism: unduly nice parents and siblings. Trends Ecol. Evol. 12:99-103.
MAYNARD SMITH, J. 1974. The theory of games and the evolution of animal conflicts. J. Theor. Biol. 47:209-221.
METCALF, R. A., AND G. S. WHITT. 1977. Intra-nest relatedness in the social wasp Polistes metricus. a genetic analysis. Behav. Ecol. Sociobiol. 2:339-351.
MIYANO, W. 1980. Life tables of colonies and workers in a paper wasp, Polistes chinensis antennalis, in central Japan (Hymenoptera:Vespidae). Res. Popul. Ecol. 25:198-209.
PAGE, R. E., AND R. A. METCALF. 1982. Multiple mating, sperm utilization, and social evolution. Am. Nat. 119:263-281.
PEMBERTON, J. M., J. SLATE, D. R. BANCROFT, AND J. A. BARETT. 1995. Nonamplifying alleles at microsatellite loci: a caution for parentage and population studies. Mol. Ecol. 4:249-252.
POLLOCK, G. B. 1996. Kin selection, kin avoidance, and correlated strategies. Evol. Ecol. 10:29-43.
QUELLER, D. C. 1994. A method for testing kin discrimination within natural colonies of social insects. Anim. Behav. 47:569-576.
QUELLER, D. C., AND K. F. GOODNIGHT. 1989. Estimating relatedness using genetic markers. Evolution 43:258-275.
QUELLER, D.C., AND J. E. STRASSMANN. 1998. Kin selection and social insects. Bioscience 48:165-175.
QUELLER, D. C., J. E. STRASSMANN, AND C. R. HUGHES. 1993. Microsatellites and kinship. Trends Ecol. Evol. 8:285-288.
QUELLER, D. C., J. M. PETERS, C. R. SOLIS, AND J. E. STRASSMANN. 1997. Control of reproduction in social insect colonies: individual and collective relatedness preferences in the paper wasp, Polistes annularis. Behav. Ecol. Sociobiol. 40:3-16.
RATNIEKS, F. L. W. 1988. Reproductive harmony via mutual policing by workers in eusocial Hymenoptera. Am. Nat. 132:217-236.
RATNIEKS, F. L. W., AND H. K. REEVE. 1992. Conflict in single-queen Hymenopteran societies: the structure of conflict and processes that reduce conflict in advanced eusocial species. J. Theor. Biol. 158:33-65.
RATNIEKS, F. L. W., AND P. K. VISSCHER. 1989. Worker policing in the honeybee. Nature 342:796-797.
REEVE, H. K. 1991. Polistes. Pp. 99-148 in K. G. Ross and R. W. Matthews, eds. The social biology of wasps. Cornell Univ. Press, Ithaca, NY.
ROSS, K. G., AND R. W. MATTHEWS, EDS. 1991. The social biology of wasps. Cornell Univ. Press, Ithaca, NY.
RUSSINA, L. Y., O. A. VISHNYAKOVA, AND E. N. LARICHEVA. 1993. Egg-laying and aggressive behaviour in three species of Polistes wasps. Proc. Coil. Soc. Ins. 2:103-109.
STARR, C. K. 1984. Sperm competition, kinship, and sociality in the aculeate Hymenoptera. Pp. 427-464 in R. L. Smith, ed. Sperm competition and the evolution of animal mating systems. Academic Press, New York.
STRASSMANN, J. E., AND D.C. MEYER. 1983. Gerontocracy in the social wasp, Polistes exclamans. Anim. Behav. 31:431-438.
STRASSMANN, J. E., C. R. HUGHES, D. C. QUELLER, S. TURILLAZZI, R. CERVO, S. K. DAVIS, AND K. F. GOODNIGHT. 1989. Genetic relatedness in primitively eusocial wasps. Nature 342:268-269.
STRASSMANN, J. E., C. R. SOLIS, J. M. PETERS, AND D. C. QUELLER. 1996. Strategies for finding and using highly polymorphic DNA microsatellite loci for studies of genetic relatedness and pedigrees. Pp. 163-180 in J. D. Ferraris and S. R. Palumbi, eds. Molecular zoology: advances, strategies, and protocols. Wiley-Liss, New York.
STRASSMANN, J. E., K. BAREFIELD, C. R. SOLIS, C. R. HUGHES, AND D.C. QUELLER. 1997. Trinucleotide microsatellite loci for a social wasp, Polistes. Mol. Ecol. 6:97-100.
TURILLAZZI, S., AND M. J. WEST-EBERHARD, EDS. 1996. Natural history and evolution of paper wasps. Oxford Univ. Press, Oxford.
WEST-EBERHARD, M. J. 1969. Social biology of polistine wasps. Misc. Publ. Mus. Zool. Univ. Mich. 140:1-101.
-----. 1986. Dominance relations in Polistes canadensis L. a tropical social wasp. Monit. Zool. Ital. (N.S.) 20:263-281.
WILSON, E. O. 1971. The insect societies. Harvard Univ. Press, Cambridge, MA.
WOYCIECHOWSKI, M., AND A. LOMNICKI. 1987. Multiple mating of queens and the sterility of workers among eusocial Hymenoptera. J. Theor. Biol. 128:317-327.
|Printer friendly Cite/link Email Feedback|
|Author:||Arevalo, Elisabeth; Strassmann, Joan E.; Queller, David C.|
|Date:||Jun 1, 1998|
|Previous Article:||Hybridization and geographic variation in two meadow katydid contact zones.|
|Next Article:||Reproductive characteristics of the flower breeding Drosophila hibisci bock (drosophilidae) in Eastern Australia: genetic and environmental...|