Naturwissenschaften

, Volume 91, Issue 11, pp 539–543

Parasitoids and competitors influence colony-level responses in the red imported fire ant, Solenopsis invicta

Authors

    • Section of Integrative Biology and Brackenridge Field Laboratory, School of Biological SciencesUniversity of Texas at Austin
    • Department of Ecology and Evolutionary BiologyRice University
  • Elizabeth A. Kawazoe
    • Section of Integrative Biology and Brackenridge Field Laboratory, School of Biological SciencesUniversity of Texas at Austin
  • Lawrence E. Gilbert
    • Section of Integrative Biology and Brackenridge Field Laboratory, School of Biological SciencesUniversity of Texas at Austin
Short Communication

DOI: 10.1007/s00114-004-0561-5

Cite this article as:
Mehdiabadi, N.J., Kawazoe, E.A. & Gilbert, L.E. Naturwissenschaften (2004) 91: 539. doi:10.1007/s00114-004-0561-5

Abstract

Social insect colonies respond to challenges set by a variable environment by reallocating work among colony members. In many social insects, such colony-level task allocation strategies are achieved through individual decisions that produce a self-organized adapting group. We investigated colony responses to parasitoids and native ant competitors in the red imported fire ant (Solenopsis invicta). Parasitoid flies affected fire ants by decreasing the proportion of workers engaged in foraging. Competitors also altered colony-level behaviours by reducing the proportion of foraging ants and by increasing the proportion of roaming majors, whose role is colony defence. Interestingly, the presence of both parasitism and competition almost always had similar effects on task allocation in comparison to each of the biotic factors on its own. Thus, our study uniquely demonstrates that the interactive effect of both parasitism and competition is not necessarily additive, implying that these biotic factors alter colony behaviour in distinct ways. More generally, our work demonstrates the importance of studying the dynamics of species interactions in a broader context.

Introduction

Social insect colonies reallocate labour according to a continually changing environment (Robinson 1992). For example, the discovery of a large ephemeral food source might require a sudden increase in foraging. This additional work force might come from reserves of unemployed workers, as in honeybees (Seeley 1982) or from workers engaged in other tasks, as in harvester ants (Gordon 1989; Gordon and Mehdiabadi 1999).

Parasites and competitors of eusocial insects also have the potential to alter a colony’s environment. For instance, previous work has shown in termites and leafcutter ants, respectively, that colonies can possibly cope with the biotic stress of disease by increasing grooming (Rosengaus et al. 1998) or by supplying additional labour to waste-removal tasks (Hart et al. 2002). Studies in other social insects have also demonstrated that interspecific interactions can influence the proportion of workers engaged in exterior tasks (Sanders and Gordon 2000). Here, we examine the potential for a phorid fly parasitoid and a native ant competitor to induce a colony-level response in the red imported fire ant (Solenopsis invicta).

Phorid fly parasitoids belong to the dipteran family Phoridae (Disney 1994). About 25 species in the genus Pseudacteon are specialized parasitoids of fire ants. Pseudacteon females oviposit into fire ants working outside the nest. The host dies several days after oviposition, when the third instar causes decapitation of the ant’s head, the site of adult fly emergence (Porter et al. 1995).

It has been suggested that these flies affect fire ants less through host mortality than through their indirect effects on foraging and interspecific interactions (e.g. Feener 1981; Orr et al. 1995; Morrison 1999). Previous work in the laboratory has shown that the phorid fly P. tricuspis can harm fire ant colonies by preferentially attacking the large-sized workers (Mehdiabadi and Gilbert 2002) that specialize in foraging and defence (Wilson 1978). Consequently, reductions in colony growth were likely caused by a combination of phorid-induced mortality and reduced colony food intake. Here we investigate whether or not colonies modify task allocation in response to parasitoids and competitors.

Materials and methods

Experimental design

We examined the independent and interactive effects of parasitism and competition on task allocation in the red imported ant using a fully factorial design. We exposed laboratory colonies to the parasitoid fly Pseudacteon tricuspis in the parasitism treatments and to the native ant Forelius mccooki in the competition treatments.

Colony collection and maintenance

This experiment had 12 replicates. We collected 12 multiple-queen colonies of fire ants and 13 multiple-queen colonies of F. mccooki in Austin, Texas, between April and August 2000. We chose this dolichoderine as the interspecific competitor because it is one of several native ants in Texas not reduced in overall range in fire ant invasion zones (Camilo and Phillips 1990).

We divided each of the 12 fire ant field colonies into four subcolonies, randomly assigning each to one of the four treatments, for a total of 48 subcolonies. We used this block design because workers from different colonies may vary in task capabilities (Porter and Tschinkel 1985). We also divided F. mccooki field colonies into 24 subcolonies. Each subcolony of both species consisted of one mated queen, roughly 5,000 workers (4,250 minors and 750 majors for fire ants; there are no worker castes in F. mccooki) and brood. We randomized colonies in their spatial arrangement in the laboratory and reared subcolonies of both species as described in Mehdiabadi and Gilbert (2002).

Observations

We placed one freeze-killed cricket and a sugar water test tube on white index cards (8 cm2) in each observation arena for two days per week over a total of 50 days, giving 14 feeding periods in total. Each of these foraging periods lasted 2 h. At the beginning of this 2-h period, we introduced four female and two male phorids into the observation arenas of each of the 12 subcolonies with parasitism and removed the barrier between the two ant species in the 12 subcolonies with competition. In the 12 subcolonies exposed to both parasitism and competition, we applied both of the above treatments. The 12 control subcolonies were untreated.

At the end of each of the 14 feeding periods, we recorded the number of ants engaged in the following tasks: (a) foraging – a worker with food in its mandibles or on the white card where food was placed, (b) defence – a worker fighting with an interspecific competitor, (c) midden work – a worker on the midden (refuse pile) or carrying refuse, such as a dead ant, in its mandibles, and (d) roaming – a worker exploring outside the nest, but not performing one of the above tasks. In addition, we distinguished the caste recruited to perform these tasks.

Statistical analyses

We conducted all statistical analyses in SAS (SAS Institute 2000), and used a repeated-measures mixed model ANOVA to examine the effects of time, competition and parasitism on the abundance and proportions of minors and majors engaged in the various tasks. We calculated proportions for each subcolony and for each caste by dividing the number of ants performing a given task by the total number of ants engaged in all exterior tasks. We performed ANOVA tests for these variables using PROC MIXED with observation (14 feeding periods), competition, and presence of phorids treated as fixed effects, and field colony (12 field-collected colonies of fire ants) and interactions of field colony with the other effects treated as random effects. We used a repeated measures first-order autoregressive covariance structure. Subcolony was the subject.

Results

Only interspecific competition significantly reduced the total number of ants engaged in tasks outside the nest over time (Fig. 1). This observed trend was due in large part to the native ant competitors altering recruitment of minors, not majors. Nevertheless, both competition and phorids had a significant effect on the mean abundance of all majors recruited to outside tasks, increasing their abundance relative to controls (F1,11=5.12, P=0.0448).
Fig. 1

Total number of ants engaged in outside tasks over the 14 foraging observation periods (±SE). Majors + minors: (i) phorids × observation day: F13,572=0.67, P=0.7925; (ii) competition × observation day: F13,572=2.94, P=0.0004; (iii) phorids × competition × observation day: F13,572=0.62, P=0.8349

Even though phorids did not affect the total number of ants performing exterior tasks, phorids did disrupt certain tasks. Phorids significantly decreased both the number (Table 1) and proportion (F1,11=5.00, P=0.047) of minors, but not majors (Table 1; proportion: F1,11=3.84, P=0.076), engaged in foraging (Fig. 2). Interspecific competitors, on the other hand, reduced proportions of both minors (F1,11=35.71, P<0.0001) and majors (F1,11=17.50, P=0.002) engaged in this task, but only decreased the number of minors performing this task relative to controls (Table 1). Phorids caused a 5% decline in the mean proportion of foraging minors, yet competitors diminished foraging of minors and majors by 26% and 23%, respectively. There was no significant interaction between biotic factors for proportional data (Fig. 2), but both phorids and competitors significantly reduced the number of foraging majors almost twofold (Table 1).
Table 1

Number of (a) minors and (b) majors engaged in outside tasks (Mean value + SE). We square-root-transformed the response variables (i.e. counts of ants) to normalize residuals, and present back-transformed least square means and standard errors. Effects of treatments on the number of (c) minors and (d) majors engaged in outside tasks (ANOVA, df=1,11 for top half and df=13,572 for bottom half). Significant values are in bold print

Treatment

Foraginga

Roaming

Midden work

Defence

All outside tasks

a

Control

72.82 (+10.11)

102 (+19)

0

176.19 (+24.88)

With phorids

66.52 (+9.67)

108 (+19)

0.02 (+0.01)

175.29 (+23.25)

With competitors

34.10 (+7.02)

106 (+19)

0

2.22 (+1.04)

153.17 (+24.82)

With phorids and competitors

22.29 (+5.74)

91 (+18)

0.01 (+0.01)

2.75 (+1.15)

128.10 (+21.33)

b

Control

1.85 (+0.66)

1.74 (+0.81)

0

4.16 (+1.46)

With phorids

1.80 (+0.65)

2.30 (+0.92)

0

4.75 (+1.55)

With competitors

1.89 (+0.67)

4.22 (+1.21)

0

0.03 (+0.02)

7.14 (+1.87)

With phorids and competitors

0.96 (+0.49)

3.56 (+1.12)

0

0.03 (+0.02)

5.46 (+1.65)

c

Source

Foraginga

Roaming

Midden work

Defence

All outside tasks

F

P

F

P

F

P

F

P

F

P

Phorids (P)

5.44

0.0397

0.31

0.5902

0.28

0.6067

1.90

0.1960

Competition (C)

30.19

0.0002

0.21

0.6551

3.42

0.0914

P×C

1.23

0.2903

1.43

0.2570

1.67

0.2229

Observation day

4.23

<0.0001

4.16

<0.0001

1.40

0.1571

5.08

<0.0001

P×Observation day

0.34

0.9858

0.72

0.7403

0.56

0.8825

0.62

0.8415

C×Observation day

2.94

0.0004

2.72

0.0010

2.97

0.0003

P×C×Observation day

0.81

0.6505

0.47

0.9427

0.54

0.9002

d

Phorids (P)

3.39

0.0928

0.01

0.9106

0.02

0.8805

0.73

0.4110

Competition (C)

1.56

0.2383

9.13

0.0116

3.38

0.0933

P×C

5.46

0.0394

3.89

0.0743

5.12

0.0448

Observation day

2.18

0.0094

2.11

0.0124

2.44

0.0038

2.38

0.0041

P×Observation day

1.54

0.0991

1.01

0.4369

0.33

0.9866

1.03

0.4221

C×Observation day

1.09

0.3660

1.84

0.0346

1.02

0.4292

P×C×Observation day

0.67

0.7946

1.46

0.1261

1.57

0.0891

a For foraging, we combined data from the two food baits before performing statistical tests

Fig. 2

Mean proportion (±SE) of minors and majors engaged in foraging, roaming, midden work and defence averaged over the 14 foraging observation periods for groups from controls (a), with phorids (b), with competitors (c), and with phorids and competitors (d). Proportions were calculated for each of the two castes by dividing the number of ants of a given caste performing a particular task by the total number of ants of that caste engaged in all outside tasks. Colonies from the phorid treatment and controls did not have an interspecific competitor, and thus, did not engage in defence. The sum of proportions across the three or four tasks for each treatment might not necessarily equal one because, during some observations, several colonies did not recruit any ants to exterior tasks

Neither phorids nor competitors significantly altered the numbers of ants recruited to midden work, and phorids did not modify the numbers of workers recruited to defence, though very few ants performed either of these tasks (Table 1). Competitors affected both the mean number (Table 1) and proportion of majors that were roaming outside the nest (F1,11=6.42, P=0.028); however, phorids did not (F1,11=0.19, P=0.67). F. mccooki caused a 20% increase in the mean proportion of fire ant majors engaged in this task relative to controls.

Discussion

This experiment reveals that phorid fly parasitism and interspecific competition influence task allocation, yet the presence of both biotic treatments almost always did not result in an additive effect.

Parasitoids decreased the proportion of workers engaged in foraging, but not those engaged in the other tasks. Phorids, in addition to native ant competitors, had little influence on midden work. The upkeep of the refuse pile did not seem to be a high-priority task for fire ants. Colonies performed midden work mainly to remove dead ants, placing them in small piles at the corners of the observation arenas. Midden work might become more important to colonies in areas infested by many phorids, because fire ants move dead ant hosts to external middens.

Phorids also had no significant effect on the proportions of fire ants engaged in defence. Furthermore, F. mccooki did not significantly increase its level of recruitment to defence in the presence of fire ant-specialized parasitoids (Mehdiabadi et al. 2004).

It seems that competitors had stronger impacts on task behaviours than parasitoids had. Competitors decreased the proportion of minors and majors engaged in foraging, but also increased the proportion of roaming majors (Fig. 2). The proportion of these majors increased in competition treatments, possibly because majors specialize in defence (Wilson 1978), implying that such a colony-level response might be adaptive. We can only speculate that the additional labour might have come from recruiting majors from inside the nest, yet these data were not statistically significant (Table 1).

A likely hypothesis as to why phorids did not increase proportions of roaming majors is that these phorids preferentially attack this size class (Morrison et al. 1997). Furthermore, Fig. 2 suggests that phorids have the potential to reduce the proportion of foraging majors, although the effect was not statistically significant. The proportional difference in foraging majors between controls and those in the phorid treatment was actually larger (0.07) than that in minors (0.04), yet we did not detect a significant effect in majors, possibly due to a small sample size. The intensity of parasitism in our study was on average one attacking fly per 200 foraging ants (Mehdiabadi and Gilbert 2002), which seems to be a realistic level according to what red imported fire ant colonies experience in their native range (Orr et al. 1995, Folgarait and Gilbert 1999). Nevertheless, in its native South America, S. invicta is exposed to several phorid species that preferentially attack different worker castes (Morrison et al. 1997); thus, the combined threat of multiple phorid species that attack both worker castes might result in stronger impacts on task allocation.

Our results support the conclusions of other studies that parasitoid flies reduce foraging by their ant hosts (e.g. Feener 1981) and, furthermore, demonstrate that phorids can influence colony-level responses by reducing the proportion of workers engaged in foraging relative to those performing other tasks. More generally, this work reveals that parasitoids and competitors are capable of altering task allocation, and that when both biotic factors are present simultaneously, as is most often the case in nature, their effects on social insects are not necessarily additive. A probable reason why we did not always detect an additive effect is that these two biotic factors alter colony behaviour in disparate ways. As shown here, competitors increased the recruitment of majors; however, phorids did not. In addition, previous work has shown that this phorid species preferentially attacks majors (Morrison et al. 1997), suggesting that their recruitment should decline. Further work is necessary to understand why these biotic factors do not always alter task allocation in the same way, and how task allocation behaviours change under more complex combinations of competition and predation in nature.

Acknowledgements

We thank S. DeWalt for statistical help; J. Crutchfield for logistical support; W. Bensen for assistance with fly stocks; C. Papp and others for rearing flies. Comments by K. Foster, T. Platt, J. Strassmann and three anonymous reviewers greatly improved this manuscript. This work was funded by the Texas Imported Fire Ant Research and Management Project to L.E.G., and a Dorothea Bennett Memorial Graduate Research Fellowship to N.J.M.

Copyright information

© Springer-Verlag 2004