Behavioral Ecology Vol. 14 No. 4: 546-553
© 2003 International Society for Behavioral Ecology
Advantages and disadvantages of large colony size in a halictid bee: the queen's perspective
Theodor-Boveri-Institut for Biosciences, University of Würzburg, Am Hubland, D-97074 Würzburg, Germany
Address correspondence to E. Strohm. E-mail: strohm{at}biozentrum.uni-wuerzburg.de.
Received 14 December 2001; revised 16 September 2002; accepted 3 October 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The size of the group of social species might influence basic aspects of productivity and social interactions. In many primitively social insects, foundress queens are basically in control of the number of workers in their first brood. We examined factors that might influence the optimal number of workers a queen should produce during the solitary founding phase in Lasioglossum malachurum (Hymenoptera, Halictidae). A priori, it seems plausible that she should produce as many workers as possible (1) to maximize colony productivity and (2) to minimize the impact of brood parasitoids. However, there might also be unfavorable consequences of a large colony size from the queen's perspective. First, the queen might incur disproportionately high costs that decrease her potential for subsequent reproduction. Second, the queen might not be able to suppress the development of ovaries in a large number of workers. As a clear advantage of a large colony size, we found an increased production of sexuals. Contrary to our expectation, in the first worker phase, nests that were parasitized by Sphecodes bees had more workers than did unparasitized nests. We found no evidence that the production of the first worker brood entailed costs to the queen. However, the degree of development of worker ovaries increased with colony size, and some degree of development was detectable with as few as four workers. This study shows that the number of workers a queen produces might depend on the interaction of several factors, some of which have not been considered in detail yet.
Key words: cost of reproduction, group size, ovarian development, parasitism, productivity, trade-off.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The size of a social group is one of the most important factors affecting the success of group members on any level of social organization (Emlen, 1991
; Wilson, 1975). The optimal group size is hypothesized to balance opposing selection pressures (Wittenberger, 1981). This hypothesis is supported by many studies on vertebrates and social spiders (e.g., Clutton-Brock et al., 1999; Uetz and Hieber, 1997; Wiklund and Andersson, 1994). In a social spider, large group size has a positive effect on mean offspring survival but a negative effect on the incidence of egg parasitism and on the probability that a female will reproduce at all (Avilés and Tufiño, 1998).
In social insects, however, the significance of group size has not received as much attention (see Crozier and Pamilo, 1996). Recently, it has been suggested that colony size per se has considerable consequences for, e.g., the conflict over reproduction, morphological skew between queens and workers, colony efficiency, complexity of social interactions, and insurance-based direct fitness benefits (Anderson and Ratnieks, 1999; Bourke, 1999; Karsai and Wenzel, 1998; Shreeves and Field, 2002). These studies examined the importance of colony size primarily across different species. However, intraspecific variation in colony size might also affect social interactions and life-history traits (Field et al., 1999, 2000), as well as colony survival (Hogendoorn and Zammit, 2001; Strassmann et al., 1988). In this study, we tested the hypothesis that colony size has an effect on aspects of social interactions and colony success in the halictid bee, Lasioglossum (Evylaeus) malachurum (Kirby 1802). Because in halictids, the number of workers in the first brood is—besides exogenous factors—under the control of the queen, we analyzed a number of factors that might select for an increase or decrease of colony size from the queen's point of view.
In the eusocial species of the genus Lasioglossum, a hibernated foundress usually raises a first brood that consists of workers only. These workers provision a second brood that consists of either sexuals or new workers, in which case the brood cycle continues. In L. malachurum, as well as in some other eusocial halictids, more worker generations are produced in warmer climates (Knerer, 1992; Miyanaga et al., 1999; Richards, 2000; Sakagami and Munakata, 1972). A priori, it appears to be advantageous for the foundress to produce as many workers in the first brood as possible because this might increase the number of sexuals produced (Michener, 1990; Oster and Wilson, 1978) and reduce parasitism (Abrams and Eickwort, 1981; Lin, 1964; Wcislo, 1997a) as well as predation (Strassmann et al., 1988; but see Shakarad and Gadagkar, 1995).
However, the problem of optimal colony size might not be trivial because, from the queen's perspective, a large number of workers might also have negative effects. First, it has been suggested that in L. malachurum, the risk of nest usurpation by floater females during the period of solitary foraging by the queen increases with the time a foundress forages by itself. The fact that foundresses stop provisioning at a certain time and close their nests and, thus, obviously do not produce the maximum number of workers possible has been interpreted as a means to counter the risk of usurpation (Kaitala et al., 1990; Smith and Weller, 1989). Second, the production of a large number of workers might entail costs (sensu Trivers, 1972) to the queen that disproportionately reduce her ability to reproduce in the future (Cant and Johnstone, 1999). The most severe form of such costs would be dying during foraging (e.g., owing to predation; Field et al., 2000; Ward and Kukuk, 1998). Third, the queen might not be able to suppress the ovarian development of a large number of workers. This is probably the case in species in which queen control is accomplished by aggression toward workers, as in halictids (for review, see Knerer, 1992; Michener, 1990; Richards, 2000). As a consequence, workers might reproduce and some resources might be allocated to the queen's grandchildren to which she is less closely related than to her own offspring. Even if workers with developed ovaries do not actually oviposit, they might allocate some resources to ovarian development (see Inglesfield and Begon, 1983; Wheeler, 1996; Wheeler and Buck, 1996) to the detriment of colony productivity.
We hypothesized an increase in colony productivity, as well as a decrease in parasitism, as possible advantages of a large number of workers in the first brood and trade-offs between the production of workers and subsequent production of sexuals, as well as an incomplete suppression of ovarian development in workers as possible disadvantages in L. malachurum.
One probable advantage of large colony size is the often reported increase of colony productivity with worker number (Lee and Winston, 1987; for review, see Michener, 1990; see also Shakarad and Gadagkar, 1995; Shreeves and Field, 2002). We predicted that large colonies are more likely to have more worker broods, to have a higher probability of producing sexuals, and to produce more sexuals. We also analyzed whether the timing and duration of the first and second worker phases have an effect on productivity.
Brood parasitoids exert an important threat to social insect colonies (see Côté and Poulin, 1995; Wcislo, 1997a,b; Wilson, 1971). A large number of workers might be able to reduce parasitism, in particular by those brood parasitoids that aggressively break into nests to oviposit in a brood cell. There is some evidence that, in some cases, guarding of the nest decreases parasitism (Abrams and Eickwort, 1981; Lin, 1964; McCorquodale, 1989; Wcislo, 1997a). However, there is apparently little information on the effect of colony size on the occurrence or rate of brood parasitism (Schmid-Hempel, 1998; Wcislo, 1997a). We predicted that the incidence of brood parasitism should decrease as a function of worker number in the study species.
The queen might suffer from the production of a large number of workers by a decreased ability for subsequent reproduction, e.g., due to the depletion of certain resources. Costs of reproduction and the resulting trade-offs between current and future reproduction have only rarely been studied in Hymenoptera (Strohm and Marliani, 2001). There seems to be the implicit assumption that queens of social insects have little or no cost of reproduction because they are amply supplied with resources by their workers. Although this might be true for queens of many highly evolved ant, wasp, and termite species, queens of halictid bees might be less well adapted to oviposit at very high rates (Knerer, 1992; with special reference to L. malachurum). This problem is probably exaggerated by the relatively large eggs of halictid bees compared with eggs of the highly eusocial species. In fact, L. malachurum might be one of the most suitable species to study possible trade-offs because its closest relatives are solitary or have only one worker brood (Danforth, 1999). In the study species, high investment in the first worker brood might therefore cause a reduced life span or a subsequently reduced availability of eggs. Evidence for a limited availability of resources would be provided by a negative correlation between size and number of workers in the first brood. A limited egg availability should manifest itself as a negative correlation between the number of workers in the first brood and the number of sexuals produced, and between the number of sexuals in successive broods.
In halictid bees, suppression of ovarian development in workers seems to be mainly caused by aggression of the queen toward the worker with the most developed ovaries (Buckle, 1985; Greenberg and Buckle, 1981; Knerer, 1992; Pabalan et al., 2000). Michener (1990) hypothesized that this mechanism might become less effective as colony size increases and supported this notion with the observation that, in smaller colonies, fewer workers are mated compared to large colonies. Also, consistent with this hypothesis, Knerer (1992) reported that, at least in rather large colonies of L. malachurum in relatively warm climates, a considerable proportion of males is produced by workers. These studies suggest that queen control might become inefficient only at rather large colony sizes. However, the actual relationship between colony size and ovarian development in workers has not been examined directly. We predicted that the incidence of mated workers and/or the degree of ovarian development of workers of the first brood should increase as a function of colony size.
|
MATERIAL AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Study species and site
Details of the nesting biology of L. malachurum are given in Knerer (1992)
and Richards (2000). We investigated a nesting aggregation of L. malachurum on a sparsely vegetated area near the Biocenter of the University of Würzburg in 1998. The aggregation consisted of about 400–500 nests, on an area of about 15 x 2 m, that were established in mid-April with a nest density of about 15 nests/m2. Focal colonies were randomly chosen early in the founding period and marked with numbered tags. Nesting activity of focal colonies (N = 51) was recorded by daily checking for burrowing and foraging activity. In the following, the members of the different broods are abbreviated by W for workers; S for sexuals; and B1, B2, or B3 for the respective number of the brood. Details of the colony cycle of the study population will be presented elsewhere (Strohm E and Bordon A, in preparation). Briefly, there was considerable variation in the nesting cycle (number of worker broods) even among nests of the study site. Foundresses provisioned a first brood (that consisted of workers only, WB1; emergence, about 5–15 June). These workers provisioned a second brood (emergence, 4–28 July) that consisted of either sexuals (SB2) only, workers only (WB2), or both workers and sexuals. In nests with new workers (WB2), a third brood of only sexuals was provisioned (SB3; emergence, 28 July–2 September). There were at least a few days without foraging activity between the provisioning phase and the emergence of the next brood. Workers of the first and the second brood partly overlapped but could be easily distinguished because all first brood workers had been color-marked (see below).
Fresh mass of workers and sexuals
To test for a negative correlation between worker size and number and to correct for differences in size between male and female sexuals for the estimation of total productivity, we determined the fresh body mass of a sample of workers and sexuals. Newly emerged individuals were caught (see below), and their fresh mass was determined to the nearest 0.1 mg on a balance (Mettler AE 160). Total production of sexuals was calculated as the number of males times the mean male mass plus the number of females times the mean female mass.
Productivity
The development of the focal colonies was observed from nest founding in mid April until bees were not active anymore at the beginning of September. Measures of queens of focal nests were not taken to avoid disturbance of the foundress and risk of abandonment or usurpation by floater females. Workers of the first and second brood (WB1 and WB2) were marked at the beginning of the flight period of the respective generation. Bees were caught by placing transparent polystyrene cups over the nest entrance. Because this was done from the beginning of the emergence period of the workers on, probably all workers had left the nest at least once. Thus, probably all workers were encountered, and their number could be determined. Marking consisted of up to three dots of acrylic paint (Robinson color, Wacofin, Fulda, Germany, or Revell Color, Revell, Germany) or paint markers (no. 751, Faber Castell, Germany) on the thorax. Catching and marking took mostly less than 2 min. Thus, disturbance of provisioning was minimized.
Determination of the number of sexuals produced was more complicated because they usually did not return to the nests after emergence, and thus, they had to be counted when they emerged. We sampled sexuals by placing traps (see above) over the nest entrances during emergence periods. Newly emerged workers and queens could be distinguished by their considerable difference in size and whether they returned to the nest with pollen. Small pollen gathering females were classified as workers. Large females that disappeared or did not gather pollen were classified as new queens. Each emerged female could be unequivocally assigned to one group. Because emergence and provisioning overlapped, particularly during the emergence period of the second brood (WB2, SB2), it was not possible to place traps over the nest entrances permanently. Thus, only those sexuals were counted that emerged during observation periods of the nests (2–8 h/day, depending on weather conditions; traps were removed if provisioning workers left or returned), and some individuals escaped recording. Therefore, the absolute number of emerging sexuals could not be determined precisely.
To nevertheless compare productivity of the WB1 and WB2 worker phases, we estimated the total number of sexuals by taking into account the daily emergence pattern of the bees. Male emergence peaked at 0800 h, and more than 90% of males emerged from 0730–1030 h. Females had a less pronounced emergence peak, at about 0830–0900 h, and about 90% emerged from 0730–1130 h. We calculated a best-fit regression curve that described the emergence in the course of the day and, by extrapolation, estimated the number of emerged individuals when no direct observations were available. For example, if a nest was not directly observed from 1000–1100 h, we assumed that the emergence pattern of this nest followed the overall pattern, and the respective proportion that had probably emerged during the period, which was not directly observed, was calculated.
Because sampling of the sexuals was particularly difficult during the emergence period of SB2, only about 60% of the estimated total number of individuals were caught and directly counted. Because the periods of sampling of emerging individuals were similar among the focal nests, there was a strong correlation between the counts obtained directly and those that were estimated (r =.89, n = 51, p <.001). In SB3, 90% of the estimated number of sexuals were directly counted with a still higher correlation between the two measures (r =.99, n = 27, p <.001). Because the same method and sampling effort was used for all nests, it is unlikely that the incomplete recording causes systematic biases of nest productivity at least relative to other nests. The absolute number of SB2, however, might be subject to some error. Analyses were done for both the directly measured and the estimated productivity. Because the results were consistent with regard to statistical significance and because p value differed by less than 10%, we only present the results for the estimated productivity. Because of the high correlation of both measures in SB3, the results were nearly identical anyhow.
Potential trade-offs—i.e., negative correlations, between size and number of WB1, between number of WB1 and SB2, between number of WB2 and SB2 sexuals, and between number of SB2 and SB3—were analyzed by using the data on the production of workers and sexuals of focal nests.
Parasitism
Emerging brood parasitoids were caught and counted by use of the emergence traps that were placed over the nest entrances for the assessment of nest productivity (see above). Nests were parasitized by Sphecodes monilicornis (Hymenoptera, Halictidae), the most important brood parasitoid of L. malachurum (Knerer, 1973; Legewie, 1925; Sick, 1993; Westrich, 1989). Other parasitoids were not encountered.
Ovarian development
WB1 workers emerged at the beginning of June. All workers of another group of 25 nests were color-marked and counted (see above). During the peak activity period of the first worker phase (8–22 June), one to three workers per nest (1.4 ± 0.7 per nest) were removed when they left to forage. Thus, no guard bees were sampled. To provide an index of worker size, head width of these workers were measured. Workers were then dissected under a stereomicroscope. The spermatheca and ovary of each specimen was transferred to a slide, and the presence of sperm and yellow bodies (corpora lutea; occurrence indicates preceding oviposition; see Peters, 1987; Tsuji, 1988) was checked by use of a compound microscope. The degree of ovarian development was classified into four categories with reference to the method of Michener and Wille (1961). If more than one worker was analyzed from one nest, we used the mean of the classes of the individual workers as an index of ovarian development to avoid pseudoreplication.
Analysis
Data were checked for assumptions of parametric tests by histogram plots. Appropriate transformations resulted in reasonable approximations of the data to normality. The number of workers, as well as the number and mass of sexuals, was square root–transformed (Sokal and Rohlf, 1995). Data are presented as mean ± 1 SD. We tested whether colony size and the start date, end date, and duration of the provisioning period had an effect on the production of sexuals by stepwise multiple regression analysis. We tested whether colony size had an effect on the occurrence of a second worker phase and whether a colony was successful (i.e., produced sexuals) by logistic regression analysis. We used t tests to test for differences between means of two groups. Because the index of ovarian development was on an ordinal scale, we used Spearman rank correlation to test for an association with colony size. Some data for individual nests are missing. Thus, sample sizes slightly differ between analyses. All analyses were conducted by using SPSS 10.0.
The number of WB1 workers might have direct and indirect effects on the total productivity of the nest: directly via the provisioning of SB2 sexuals and indirectly via the production of WB2 workers that in turn provision SB3 sexuals. To estimate the total effect of WB1 workers on total productivity, we constructed a path diagram and calculated path coefficients for the different paths (Schumacker and Lomax, 1996; Sokal and Rohlf, 1995) that lead to the dependent variable "total mass of sexuals." Path coefficients are standardized regression coefficients. Thus, a path coefficient of 0.5 means that if the independent variable (also called predictor variable) changes by 1 SD, the dependent variable (criterion variable) will on average change by 0.5 SD.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Body mass of workers and sexuals
Newly emerged workers of WB1 weighed 15.5 ± 2.3 mg (n = 48), those of WB2 15.1 ± 1 mg (n = 9). Newly emerged queens had a fresh mass of 23.2 ± 5.4 mg in SB2 (n = 22) and 23.5 ± 2.4 mg in SB3 (n = 79). The respective values for males were 13.3 ± 2.9 mg in SB2 (n = 26) and 14.4 ± 5 mg in SB3 (n = 151). These values were multiplied with the number of male and female sexuals to obtain the total mass of sexuals for the following analyses.
Colony size and productivity
The mean number of WB1 and WB2 workers of the focal nests was 4.25 ± 2.3 (n = 51) and 7.5 ± 8.3 (n = 27) individuals in nests that had a second worker phase, respectively. In some nests, a few WB1 were still alive when WB2 emerged and started foraging. However, the old WB1 showed strong signs of senescence (wing wear, very low level of activity) and disappeared rapidly. Thus, they did not contribute to the provisioning of the WB2 phase, and we use the number of newly emerged workers as the measure of colony size during WB2. Stepwise multiple regression analysis revealed that the total mass of SB2 sexuals was significantly influenced by the number of WB1 (r =.5, n = 50, p <.001) (Figure 1A). The duration (partial r =.03, p =.83), date of beginning (partial r = -.08, p =.59), and date of ending of the first worker phase (partial r = -.05, p =.73) had, however, no significant effect. If only colonies were considered that produced any SB2, there was a weak but not significant correlation between the number of WB1 and the mass of SB2 (r =.27, n = 40, p =.08). Thus, the effect of WB1 number on colony success was primarily owing to a significantly higher probability of producing sexuals at all (logistic regression: Wald 
2 = 13.5, df = 1, p <.001). With four WB1, the probability to produce any progeny (workers or sexuals) was about 95% (probability of success = e-2.2+2.6xWB1). The number of WB2 increased significantly with the number of WB1 (r =.39, n = 50, p =.005; equation: 
= -0.17 + 0.78 x 
). The correlation is still significant if only nests with a second worker phase were considered (r =.39, n = 27, p =.04). The logistic regression of number of WB1 on the existence of a second worker phase was, however, not significant (Wald 2 = 1.5, df = 1, p =.22). Thus, contrary to the production of sexuals, the number of WB2 increased more or less continuously with the number of WB1.
|
The mass of SB3 sexuals depended significantly on the number of WB2 (stepwise multiple regression: r =.8, n = 27, p <.001) (Figure 1B) but not on the duration (partial r =.15, p =.32), date of beginning (partial r = -.056, p =.72), or date of ending (partial r =.056, p =.72) of the WB2 phase. Whether a colony successfully produced sexuals depended weakly but not significantly on the number of WB2 (logistic regression: Wald
2 = 3.39, df = 1, p =.066). In contrast to the WB1 worker phase, the number of WB2 had still a highly significant positive effect on the mass of SB3, if only successful colonies were considered (r =.9, N = 14, p <.001; equation: 
= -1.7 + 6.3 x 
). Thus, the number of WB1 primarily determined whether a colony was successful in producing SB2 at all, whereas the number of WB2 determined the quantitative output of SB3. To encompass an indirect effect of the number of WB1 on total success mediated by the correlation between worker number in WB1 and WB2, we conducted a path analysis to assess the contribution of WB1 and WB2 to total sexual production (Figure 2). The paths leading from the number of WB1 to the total mass of sexuals had a combined coefficient of 0.57 (see legend of Figure 2). The path coefficient of the "direct" effect via production of SB2 was 0.335. The difference, 0.235, is the coefficient for the indirect effect, via the production of WB2. The total path coefficient for WB2 worker number was 0.48 and, thus, smaller than the coefficient for WB1. This means that an increase in the number of WB1 of 1 SD on average increased production of sexuals by 0.57 SD, whereas an increase of 1 SD in the number of WB2 increased production of sexuals by 0.48 SD. Because SD of worker numbers in WB1 and WB2 differed, the actual increase in the mass of sexuals produced for one additional worker is on average 69 mg in WB1 (this equals about three queens or five males) and 38 mg in WB2 (1.6 queens or 2.7 males). Anyhow, this result suggests that a queen would on average benefit from producing more WB1 workers by having a larger number of progeny.
|
S total: [0.5 x 0.67] + [0.08 x 0.63] + [0.39 x 0.76 x 0.63] = 0.57)Colony size and parasitism
S. monilicornis emerged among the second brood from 12 of the 51 focal nests (in seven nests one parasitoid, in five nests two parasitoids) and among the third brood from five of the 27 nests with a WB2 worker phase (in four nests one parasitoid, in one nest two parasitoids). The mean number of WB1 workers of nests that had been parasitized during the WB1 phase (5.2 ± 2.4, n = 12) was slightly but significantly higher than the number of WB1 of unparasitized nests (3.9 ± 1, n = 36; t test: t = 2.2, p =.044). Colony productivity might confound this analysis because in highly productive nests more brood cells are exposed to parasitism. Productivity, however, did not differ between parasitized and unparasitized nests (t = 0.075, p =.94). Thus, contrary to our prediction, large colonies had a higher probability of being parasitized during the WB1 phase. For the WB2 phase, there was no significant difference in the mean number of workers of parasitized (6.5 ± 3.1, n = 5) and unparasitized nests (7.7 ± 6.5, n = 22; t = 0.7, p =.5). Also, productivity did not differ between the two groups of nests (t = 0.67, p =.51). However, the sample size of parasitized nests was too low to obtain a conclusive result.
Trade-offs in the production of progeny
The mean fresh mass of WB1 was not significantly correlated with the respective number of WB1 in a nest (Pearson correlation: r = -.1, n = 48, p =.5). The number of WB1 was positively correlated with colony productivity of this phase (SB2 and WB2 progeny, see above). There was no indication of a trade-off between the production of WB2 and SB2 (r =.08, n = 51, p =.56), even when we controlled for the number of WB1 as a potential indicator of resource availability of a colony (partial r = -.06, p =.87). There was also no negative correlation between the mass of SB2 and SB3 (r =.13, n = 27, p =.36), even when we controlled for the number of WB1 and WB2 workers, which might be an indicator of resource availability of a colony (partial r = -.07, p =.64). There was even a tendency that nests with WB2 produced more sexuals in the first brood (282 ± 64 mg, n = 27) than did nests that ceased provisioning after the first worker phase (142 ± 79 mg, n = 24; t test: t = 2.0, p =.047). Thus, there is no indication of trade-offs between the production of WB1 workers and sexuals or sexuals in successive broods.
Colony size and development of worker ovaries
The nests sampled for the assessment of ovarian development contained 5.8 ± 2.2 WB1 (range, 1–11). None of the workers had sperm in the spermatheca or yellow bodies in the ovary (N = 34). Sixteen workers (47%) had developed ovaries (class 2–4); six of these (18%) had a mature egg. The degree of ovarian development was independent of the head width of the workers (Spearman rank correlation: rs = -.18, n = 32, p =.3). However, the degree of ovarian development in workers increased significantly with increasing colony size (rs =.48, n = 25, p =.014) (Figure 3).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Colony size and productivity
For both worker phases (WB1, WB2), there was a significant positive effect of worker number on productivity. In WB1, this resulted mainly from the effect that worker number had an influence on whether a colony actually produced sexuals. This suggests that a minimum number of workers is necessary to successfully rear progeny during the WB1 worker phase. The possible reasons for this are unknown. In bumblebees also a certain number of workers is necessary to rear any progeny (for review, see Michener, 1990
; Müller and Schmid-Hempel, 1992). If a certain number of workers is necessary to be successful, foundresses should adjust their effort to produce at least this minimum number of workers. The least number of workers necessary to provision a brood is probably two, one guard and one forager. However, the number of WB1 workers that is associated with a nearly 100% probability of success is about four to six. Interestingly, this coincides with the range of the actual mean number of workers in the first brood (4.2 and 5.8).
That a minimum number of workers is needed is further supported by the remarkable similarity of colony sizes of our population with four populations from different climates reported by Packer and Knerer (1985; Isle of Wight, UK: 6.7; Marseille, France: 6.8; Dordogne, France: 5.1; Estepona, Spain: 6.5) and one reported by Paxton et al. (2002; Tübingen, Germany: 4.7). The hypothesis of a certain minimum number of WB1 workers in L. malachurum does not contradict the hypothesis that foundresses cease provisioning and close the nests to avoid failure owing to usurpation (Kaitala et al., 1990) or mortality (Ward and Kukuk, 1998), but would specify the stage at which nest closing should occur, i.e., when about four to six workers have been produced.
Colony size and parasitism
For WB2, no correlation between colony size and incidence of parasitism was observed, possibly owing to the small sample size. Contrary to our prediction, during the WB1 phase larger colonies were more likely to be parasitized by S. monilicornis than were smaller colonies. Unfortunately, because of the small number of parasitoids per nest, it was only possible to test for an effect of colony size on the presence/absence of parasitism. Although unlikely, it is still possible that the probability of parasitism per brood cell decreased with colony size. Assuming our result is true, why might the probability of parasitism increase with colony size although more workers are available for defense? Possibly a nest becomes more conspicuous, both visually and olfactorily (Legewie, 1925; Sick, 1993), when more workers are active at the nest entrance. Furthermore, the guard bee might sometimes be confused when many workers are trying to enter the nest simultaneously, and a S. monilicornis female might creep into a nest. In any case, our result at least suggests that parasitism by S. monilicornis might be a factor that counters the selection for large colony size in WB1.
Trade-offs in the production of progeny
The number of workers in WB1 that are produced by the queen might be compromised by disproportionately high costs to the queen. Contrary to our expectation, we did not find evidence for such trade-offs, between the size and the number of WB1 workers, between the number of workers and sexuals, or between the number of sexuals of successive generations. However, trade-offs are difficult to detect (see Lessells, 1991; Reznick et al., 2000). If there is a large variation in the amount of resources available to individuals (or colonies), the superior individuals will be able to allocate large amounts of resources to all traits. Thus, across individuals there will be no or even a positive relationship between these traits, although within an individual these traits compete for resources and are, thus, negatively correlated (Lessells, 1991; Reznick et al., 2000; van Noordwijk and de Jong, 1986). Such differences in resource availability among colonies, which possibly mask a trade-off, might be indicated by the number of workers. However, even when we statistically controlled for worker number, we could not detect any trade-off. The result that colonies with only one worker phase had a lower output of SB2 than did colonies that had a second worker phase suggests that colonies differ in aspects that are not represented by the number of workers (e.g., "qualitative" aspects of queens and/or workers). As a consequence, we cannot conclusively rule out the existence of trade-offs and their possible role in shaping the pattern of resource allocation of the queen. Experimental manipulation (Lessells, 1991; Strohm and Marliani, 2002) of the queen's workload or resource levels is needed for a more conclusive analysis of possible trade-offs.
Colony size and development of worker ovaries
In the nests sampled for the determination of ovarian development, the mean number of WB1 workers was quite similar to the four populations studied by Packer and Knerer (1985; see above). The low proportion of mated workers of the first generation in these four populations (0–0.8%; Packer and Knerer, 1985) is consistent with the absence of mated workers in our study. In the second (and third) worker brood, more workers can be mated (up to 30%; Knerer, 1992; Richards, 2000). The overall proportion of 47% of workers with developed ovaries in L. malachurum in this study is somewhat higher than the 32% reported by Knerer and Plateau-Quenu (1967) and the 3.4%, 6.8%, 17.6%, and 30.1% reported by Packer and Knerer (1985) for four different European populations.
Our result that the proportion of workers with developed ovaries increased with colony size supports Michener's (1990) prediction that the queen's ability to inhibit workers might become less effective in larger colonies (see also Buckle, 1985; Knerer, 1992; Nakata and Tsuji, 1996; Richards, 2000). This might be owing to a dilution effect because an individual worker might receive less aggression from the queen in large colonies (see also Richards et al., 1995). Furthermore, our study suggests that queen control might already become less effective at surprisingly small colony sizes (four to six workers). Paxton et al. (2002) have recently shown that L. malachurum workers successfully reproduce in nests with a mean number of only 4.7 workers.
Michener's hypothesis explains incomplete queen control by a constraint of the queen and, thus, represents a proximate explanation. Reproductive skew models might provide additional insights on an ultimate level, although existing models do not explicitly consider colony size. Assuming that queen aggression and reproduction entails some cost on the queen (i.e., a decrement in future reproduction; sensu Trivers, 1972; see Cant and Johnstone, 1999), the decreased suppression of worker ovaries in large colonies might be interpreted as the evolutionary result of balancing of opposing effects. Given aggression towards workers is costly (Reeve et al., 1998), increased ovarian development of workers in larger colonies might be the consequence of a queen's optimal allocation of resources between its own reproduction and aggressive prevention of worker reproduction. Second, the queen could benefit from indirect fitness returns by reproduction of her daughter workers if large amounts of provisions are available, but an accordingly high rate of egg laying would be disproportionately costly. That reproduction in L. malachurum is costly is supported by the fact that in Mediterranean habitats, foundresses of L. malachurum die before the end of the nesting cycle (Richards, 2000). Furthermore, considering the short period of provisioning activity, a certain degree of ovarian development might predispose workers to rapidly adopt egg-laying status if the queen dies before the end of the season and might, thus, be interpreted as a form of insurance strategy (Field et al., 2000; Gadagkar, 1990; Shreeves and Field, 2002) of the queen.
|
ACKNOWLEDGEMENTS |
|---|
We are grateful to Andrew Bourke, Jeremy Field, Jürgen Liebig, Judith Korb, Robert Paxton, and two anonymous reviewers for very helpful comments on earlier drafts of the manuscript. This study was supported by the Deutsche Forschungsgemeinschaft (SFB 554, TPB3).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abrams J, Eickwort GC, 1981. Nest switching and guarding by the communal sweat bee Agapostemon virescens (Hymenoptera, Halictidae). Insectes Soc 28:105-116.[CrossRef]
Anderson C, Ratnieks FLW, 1999. Task partitioning in insect societies, I: effect of colony size on queuing delay and colony ergonomic efficiency. Am Nat 154:521-535.[CrossRef][Medline]
Avilés L, Tufiño P, 1998. Colony size and individual fitness in the social spider Anelosimus eximius. Am Nat 152:403-418.[CrossRef][Web of Science][Medline]
Bourke AFG, 1999. Colony size, colony complexity and reproductive conflict in social insects. J Evol Biol 12:245-257.[CrossRef][Web of Science]
Buckle GR, 1985. Increased queen-like behaviour of workers in large colonies of the sweat bee Lasioglossum zephyrum. Anim Behav 33:1275-1280.[CrossRef]
Cant MA, Johnstone RA, 1999. Costly young and reproductive skew in animal societies. Behav Ecol 10:178-184.
Clutton-Brock TH, Gaynor D, McIlrath GM, Maccoll ADC, Kansky R, Chadwick P, Manser M, Skinner JD, Brotherton PNM, 1999. Predation, group size and mortality in a cooperative mongoose, Suricata suricata. J Anim Ecol 68:672-683.[CrossRef]
Côté IM, Poulin R, 1995. Parasitism and group size in social animals: a meta-analysis. Behav Ecol 6:159-165.[Abstract]
Crozier RH, Pamilo P, 1996. Evolution of social insect colonies. Oxford: Oxford University Press.
Danforth BN, 1999. Phylogeny of the bee genus Lasioglossum (Hymenoptera: Halictidae) based on mitochondrial COI sequence data. Syst Entomol 24:377-393.[CrossRef]
Emlen ST, 1991. Evolution of cooperative breeding in birds and mammals. In: Behavioural ecology: an evolutionary approach, 3rd ed (Krebs JR, Davies NB, eds). Oxford: Blackwell; 301–337.
Field J, Shreeves G, Sumner S, 1999. Group size, queuing and helping decisions in facultatively eusocial hover wasps. Behav Ecol Sociobiol 45:378-385.[CrossRef][Web of Science]
Field J, Shreeves G, Sumner S, Casiraghi M, 2000. Insurance-based advantages to helpers in a tropical hover wasp. Nature 404:869-871.[CrossRef][Medline]
Gadagkar R, 1990. Evolution of eusociality: the advantage of assured fitness returns. Phil Trans R Soc Lond B 329:17-25.[CrossRef]
Greenberg L, Buckle GR, 1981. Inhibition of worker mating by queens in a sweat bee, Lasioglossum zephyrum. Insectes Soc 28:347-352.[CrossRef]
Hogendoorn K, Zammit J, 2001. Benefits of cooperative breeding through increased colony survival in an allodapine bee. Insectes Soc 48:392-397.[CrossRef]
Inglesfield C, Begon M, 1983. The ontogeny and cost of migration in Drosophila subobscura Collin. Biol J Linn Soc 19:9-15.
Kaitala VB, Smith H, Getz WM, 1990. Nesting strategies of primitively eusocial bees: a model of nest usurpation during the solitary state of the nesting cycle. J Theor Biol 144:445-471.
Karsai I, Wenzel JW, 1998. Productivity, individual-level and colony-level flexibility, and organization of work as a consequences of colony size. Proc Natl Acad Sci USA 95:8665-8669.
Knerer G, 1973. Periodizität und Strategie der Schmarotzer einer sozialen Schmalbiene, Evylaeus malachurus (K.) (Apoidea: Halictidae). Zool Anz 190:41-63.
Knerer G, 1992. The biology and social behaviour of Evylaeus malachurus (K.) (Hymenoptera; Halictidae) in different climatic regions of Europe. Zool Jahrb Abt Syst Oekol Geogr Tiere 119:261-290.
Knerer G, Plateaux-Quenu C, 1967. Sur la production des males chez les Halictinae (Insectes Hymenoptères) sociaux. C R Acad Sci 264:1096-1099.
Lee PC, Winston ML, 1987. Effects of reproductive timing and colony size on the survival, offspring colony size and drone production in the honey bee (Apis mellifera). Ecol Entomol 12:187-195.
Legewie H, 1925. Zum Problem des tierischen Parasitismus, I. Teil: die Lebensweise der Schmarotzerbiene Sphecodes monilicornis K. (= subquadratus) (Hymenoptera: Apoidea). Z Morph Ökol Tiere 4:430-464.[CrossRef]
Lessells CM, 1991. The evolution of life histories. In: Behavioural ecology: an evolutionary approach, 3rd ed (Krebs JR, Davies NB, eds). Oxford: Blackwell; 32–68.
Lin N, 1964. Increased parasitic pressure as a major factor in the evolution of social behavior in halictine bees. Insectes Soc 11:187-192.[CrossRef]
McCorquodale D, 1989. Nest defense in single- and multifemale nests of Cerceris antipodes (Hymenoptera: Sphecidae). J Insect Behav 2:267-276.
Michener CD, 1990. Reproduction and caste in social halictine bees. In: Social insects: an evolutionary approach to castes and reproduction (Engels W, ed). Heidelberg: Springer-Verlag; 77–122.
Michener CD, Wille A, 1961. The bionomics of a primitively social bee, Lasioglossum inconspicuum. Univ Kansas Sci Bull 42:1123-1202.
Miyanaga R, Maeta Y, Sakagami SF, 1999. Geographical variation of sociality and size-linked color patterns in Lasioglossum (Evylaeus) apristum (Vachal) in Japan (Hymenoptera, Halictidae). Insectes Soc 46:224-232.[CrossRef]
Müller CB, Schmid-Hempel P, 1992. Correlates of reproductive success among field colonies of Bombus lucorum: the importance of growth and parasites. Ecol Entomol 17:343-353.[CrossRef]
Nakata K, Tsuji K, 1996. The effect of colony size on conflict over male-production between gamergate and dominant workers in the ponerine ant Diacamma sp. Ethol Ecol Evol 8:147-156.
Oster GF, Wilson OE, 1978. Caste and ecology in the social insects. Princeton, New Jersey: Princeton University Press.
Pabalan N, Davey KG, Packer L, 2000. Escalation of aggressive interactions during staged encounters in Halictus ligatus Say (Hymenoptera: Halictidae), with a comparison of circle tube behaviors with other halictine species. J Insect Behav 13:627-650.[CrossRef]
Packer L, Knerer G, 1985. Social evolution and its correlates in bees of the subgenus Evylaeus (Hymenoptera; Halictidae). Behav Ecol Sociobiol 17:143-149.
Paxton RJ, Ayasse M, Field J, Soro A, 2002. Complex sociogenetic organization and reproductive skew in a primitively eusocial sweat bee, Lasioglossum malachurum, as revealed by microsatellites. Mol Ecol 11:2405-2416.[CrossRef][Medline]
Peters CP, 1987. The reproductive division of labor in the queenless ponerine ant Rhytidoponera sp. 12. Insectes Soc 34:75-86.
Reeve HK, Emlen ST, Keller L, 1998. Reproductive sharing in animal societies: reproductive incentives or incomplete control by dominant breeders? Behav Ecol 9:267-278.
Reznick D, Nunney L, Tessier A, 2000. Big houses, big cars, superfleas and the costs of reproduction. Trends Ecol Evol 15:421-425.[CrossRef][Medline]
Richards MH, 2000. Evidence for geographic variation in colony social organization in an obligately social sweat bee, Lasioglossum malachurum Kirby (Hymenoptera; Halictidae). Can J Zool 78:1259-1266.[CrossRef]
Richards MH, Packer L, Seger J, 1995. Unexpected patterns of parentage and relatedness in a primitively eusocial bee. Nature 379:239-241.
Sakagami SF, Munakata M, 1972. Distribution and bionomics of a transpalaearctic eusocial halictine bee, Lasioglossum calceatum, in Northern Japan, with reference to its solitary life cycle at high altitude. J Fac Sci Hokkaido Univ Ser 6 Zool 18:411-439.
Schmid-Hempel P, 1998. Parasites in social insects. Princeton, New Jersey: Princeton University.
Schumacker RE, Lomax RG, 1996. A beginner's guide to structural equation modelling. Mahwah, New Jersey: Erlbaum Associates.
Shakarad M, Gadagkar R, 1995. Colony founding in the primitively eusocial wasp, Rhopalidia marginata (Hymenoptera: Vespidae). Ecol Entomol 20:273-282.
Shreeves G, Field J, 2002. Group size and direct fitness in social queues. Am Nat 159:81-95.[CrossRef][Web of Science][Medline]
Sick M, 1993. Auffinden und olfaktorisches Erkennen von Wirtsnestern durch Kuckucksbienen (Gattung Sphecodes, Halictidae) und deren verwandtschaftliche Beziehungen zu den Wirtsbienen (PhD dissertation). Tübingen: Eberhard-Karls-Universität.
Smith BH, Weller C, 1989. Social competition among gynes in halictine bees: the influence of bee size and pheromones on behavior. J Insect Behav 2:397-411.[CrossRef]
Sokal RR, Rohlf FJ, 1995. Biometry: the principles and practice of statistics in biological research, 3rd ed. New York: Freeman.
Strassmann JE, Queller DC, Hughes CR, 1988. Predation and the evolution of sociality in the paper wasp Polistes bellicosus. Ecology 69:1497-1505.[CrossRef][Web of Science]
Strohm E, Marliani A, 2002. The cost of parental care: prey hunting in a digger wasp. Behav Ecol 13:52-58.
Trivers RL, 1972. Parental investment and sexual selection. In: Sexual selection and the descent of man (Campbell B, ed). Chicago: Aldine; 136–179.
Tsuji K, 1988. Obligate parthenogenesis and reproductive division of labor in the Japanese queenless ant Pristomyrmex pungens: comparison of intranidal and extranidal workers. Behav Ecol Sociobiol 23:247-256.[CrossRef][Web of Science]
Uetz GW, Hieber CS, 1997. Colonial web building spiders: balancing the costs and benefits of group-living. In: Social behavior in insects and arachnids (Choe JC, Crespi BJ, eds). Cambridge: Cambridge University Press; 458–475.
van Noordwijk AJ, de Jong G, 1986. Acquisition and allocation of resources: their influence on variation in life history tactics. Am Nat 128:137-142.[CrossRef][Web of Science]
Ward SA, Kukuk PF, 1998. Context-dependent behavior and the benefits of communal nesting. Am Nat 152:249-263.[CrossRef]
Wcislo WT, 1997a. Behavioral environments of sweat bees (Halictinae) in relation to variability in social organization. In: Social behavior in insects and arachnids (Choe JC, Crespi BJ, eds). Cambridge: Cambridge University Press; 316–332.
Wcislo WT, 1997b. Invasion of nests of Lasioglossum imitatum by a social parasite, Paralictus asteris (Hymenoptera: Halictidae). Ethology 103:1-11.
Westrich P, 1989. Die Wildbienen Baden-Württembergs. Stuttgart: Ulmer.
Wheeler D, 1996. The role of nourishment in oogenesis. Ann Rev Entomol 41:407-431.[CrossRef][Web of Science][Medline]
Wheeler D, Buck N, 1996. Depletion of reserves in ant queens during claustral colony founding. Insectes Soc 43:297-302.[CrossRef]
Wiklund CG, Andersson M, 1994. Natural selection of colony size in passerine bird. J Anim Ecol 63:765-774.[CrossRef]
Wilson EO, 1971. Insect societies. Cambridge, Massachusetts: Belknap.
Wilson EO, 1975. Sociobiology, the new synthesis. Cambridge, Massachusetts: Harvard University Press.
Wittenberger J, 1981. Animal social behavior. Belmont, California: Wadsworth.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||