Behavioral Ecology Advance Access originally published online on December 16, 2006
Behavioral Ecology 2007 18(2):337-344; doi:10.1093/beheco/arl089
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Mimicry in hoverflies (Diptera: Syrphidae): a field test of the competitive mimicry hypothesis
Department of Biology, Carleton University, 1125 Colonel By Drive, Ottawa, Ontario K1S 5B6, Canada
Address correspondence to T.N. Sherratt. E-mail: sherratt{at}ccs.carleton.ca.
Received 28 June 2006; revised 9 November 2006; accepted 11 November 2006.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although most studies on the evolution of mimicry and warning coloration in insects have considered predators as the major selective force, it is possible that competition for food resources could also facilitate selection for these conspicuous signals. For example, when warningly colored social wasps visit flowers, then they frequently behave aggressively toward heterospecifics, and they also attack and feed on other flying insects. Under these conditions, a resemblance to a wasp might provide a mimetic hoverfly with improved access to floral resources by reducing the frequency with which it is disturbed by other pollinators. We experimentally evaluated whether wasp-like colors and patterns were important in preventing other flower visitors from sharing the same flower resource, using pairwise presentations of both natural and artificial prey in the field. Flower visitors were more likely to visit unoccupied flowers compared with the flowers pinned with either natural or artificial specimens in 2 plant species with different inflorescences. However, flower visitors did not show a significantly reduced rate of visitation to flowers pinned with specimens bearing wasp-like colors and patterns compared with the flowers occupied by similar-sized specimens that were nonmimetic. Overall, we found no compelling evidence in this study to support the contention that wasp-like warning signals of hoverflies prevent other flower visitors from sharing flower resources, although insects showed a greater tendency to avoid visiting flowers pinned with a wasp compared with flowers pinned with a nonmimetic fly.
Key words: competition, flower visitors, hoverfly, mimicry, wasp.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The resemblance of a palatable (edible) species to an unpalatable (or otherwise defended) species is referred to as Batesian mimicry (Bates 1862
; Edmunds 1974). Although the phenomenon of mimicry is one of the most celebrated examples of the power of natural selection, there are still many outstanding questions and a recognized shortage of field studies, particularly on nonlepidopteran groups (Ruxton et al. 2004).
Many species of hoverfly (Diptera: Syrphidae) are palatable to their potential avian predators (Mostler 1935) and mimic noxious hymenopteran models (Gilbert 2005). One of the most popular explanations to date is that this resemblance helps deter predation by vertebrate (e.g., Mostler 1935; Dittrich et al. 1993; Green et al. 1999) and invertebrate predators (Kauppinen and Mappes 2003; see also Rashed et al. 2005). However, adult hoverflies frequently visit flowers for nectar and pollen (Branquart and Hemptinne 2000; Gilbert 1981); so it is conceivable that the resemblance to a wasp or bee also serves to reduce the frequency and intensity of competitive interactions on inflorescences, a phenomenon we have called "competitive mimicry." In employing signals to gain access to resources rather than to avoid predation, competitive mimicry, if substantiated, may be closer to the definition of "aggressive mimicry" (Cott 1940; Ruxton et al. 2004), than Batesian mimicry, although both forms of mimicry involve exploitation of the signal receivers (Holen and Johnstone 2004).
There are numerous examples of conspecifics or heterospecifics reducing the reproductive success of an individual either by directly interfering with the individual's ability to obtain resources or by causing it to expend energy in responding to these competitors (Schoener 1974). Several studies have shown that flower visitors, particularly social bees and wasps, actively compete with other flower visitors over resources. For example, it has been suggested that food source competition and aggressive encounters among the bee species Trigona conifrons, T. fimbriata, T. apicalis, and T. melina (Apidae, Meliponinae) has resulted in the separation of these species in time and space (Nagamitsu and Inoue 1997). Similarly, Biesmeijer et al. (1999) reported interspecific food source competition between the 2 species of stingless bees (Hymenoptera: Apidae), Melipona beecheii and Melipona fasciata, noting that M. beecheii tended to avoid landing close to M. fasciata on artificial feeders. Parrish and Fowler (1983) experimentally showed that when wasp species, Vespula maculifrons and Vespula germanica encountered one another on a food resource, their aggressiveness toward heterospecifics was higher than that directed toward conspecifics. More generally, Kevan and Baker (1983) report studies, principally from Kikuchi (1963), in which "subordinate" insect species were displaced to older and less desirable flowers.
Of course, many species of Hymenoptera, such as social vespid wasps, are active generalist predators (see Richter 2000), which raises the possibility that mimetic hoverflies could gain improved access to resources not only through their resemblance to a vigorous competitor but also to a predator. It is now well known that the behavior of insects when visiting flowers is sensitive to risk. For example, even Hymenoptera such as honeybees (Dukas 2001) and smaller sized bumblebees (Dukas and Morse 2003) prefer to visit "safe flowers" without potential predators compared with equally rewarding flowers that could pose a potential danger. Similarly, Howarth et al. (2004) noted a significant inverse relationship between the number of an imperfect wasp mimic, Helophilus pendulus, and the number of wasps observed and suggested that these hoverflies avoid flowers with wasps on them to minimize their predation risk.
Although Hymenoptera are known to be competitive flower foragers, we are not aware of any test of whether the mimicry of a noxious hymenopteran model provides any form of competitive advantage. Nickol (1994) treated flower-visiting hymenoptera as competitors for the hoverfly, Volucella zonaria, and considered them as possible receivers for the mimicry signals, but his theory remains untested. If wasp mimicry provides any form of competitive advantage, then one might expect that pollinators occupying the same inflorescence would come into physical contact with mimics less frequently than with similar-sized specimens that were not warningly colored. Alternatively, or in addition, flowers containing a warningly colored mimic should receive fewer visitors than unoccupied flowers and flowers with nonmimetic specimens. We set out to test these predictions using replicated field experiments.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
We experimentally investigated the rate of visitation of insects and their subsequent behavior, when landing on inflorescences of both goldenrod (Solidago canadensis, Asteraceae [Compositae]) and wild carrot (Daucus carota, Apiaceae [Umbelliferae]).
Time and location of study
Our study was conducted at Ottawa's Central Experimental Farm in 2 connected open fields near Carleton University, Canada (44°34'N, 79°15'W). Experiments on Goldenrod flowers were performed, approximately sequentially, from 10 August to 20 September in both 2004 and 2005. Due to earlier flowering times, experiments on wild carrot flowers were conducted from 22 June to 5 August 2005. All experiments took place on warm (temperature range 27–33 °C) sunny days between 10 AM and 1 PM local time. We also conducted weekly surveys to identify other insect species present in our experimental fields. Each survey involved an experimenter crossing the fields 4 times in a zigzag manner, for approximately 4–5 min, while swiping a hand-held net into the inflorescences. This allowed us to record the presence and estimate the relative abundance of the potential flower visitors in our study area.
Overall design
Each replicate consisted of simultaneously observing 2 randomly chosen flowers of the same species in the same condition (same size, same exposure to pollinators, etc.), and at least 1 m apart, in the field for 10 consecutive minutes. All observations were made from at least a 2-m distance to minimize the effect of the experimenter's presence on flower visitations. Depending on the design of each experiment, either both of the chosen flowers were pinned with a single specimen (such as a mimetic hoverfly and a nonmimetic fly) or only one of them (chosen at random) was pinned with a specimen and the other one was left unoccupied (see below). A variety of events, including the number of visits by different species (landing on the flower), the timing of visits, the length of time visitors spent on each flower, and any behavioral interactions were recorded. We also recorded the number of investigations not resulting in physical contact with the flowers ("investigations" were defined as independent inspections made by flower visitors within 10 cm—repeated movements by the same individual closer to the flower were treated as a single inspection). The taxonomic level to which visiting insects were ascribed varied between taxonomic groups, but in all cases it was to at least family, and often genus.
Pinned specimens
Both natural and artificial specimens were pinned out onto flowers in our experiments. All natural prey items were used only once, so that each replicate involved a new specimen. All of our natural specimens had been killed, dried, and kept in the fridge (at 4 °C, with no noticeable change in color intensity or tone) for at least 4 weeks prior to each experiment to reduce the potential odor effect on the experimental results (see Gamboa et al. 1996). All species used in the study had a similar total body size range (measured from forehead to the tip of the last visible segment of the abdomen, they were all 10–13 mm in length), although their shapes differed. Our artificial specimens consisted of 2 ellipsoid plastic beads, each 3 x 5 mm (diameter x length). We painted the 2 beads either black or black-and-yellow striped and attached them using a pin as a central axis. We chose to use the above dimensions and shape so as to resemble the general shape of a live flower-visiting insect.
Water-based colors (Crafter's Acrylic, DecoArt, Stanford, KY) were used to paint artificial prey types and natural prey items as required. Mock-painted hoverflies, Spilomyia longicornis, received a similar amount of paint as individuals in their comparison groups but were painted so that they retained much of their former appearance (yellow stripes and patches on the thorax and abdomen). To make the hoverflies black, we painted the yellow stripes with black color. Although the paints used were not different from the natural coloration of wasps to our eyes, we measured the reflectance of both the black-and-yellow paints using an Ocean Optics USB2000 spectrometer (Ocean Optics, Dunedin, FL) and a 200-micron reflection probe at a 45° angle from the sample surface. Illumination was provided via a pulsed xenon (PX-2) light source. Spectra were recorded at 1-nm intervals from 350 to 700 nm and measured relative to a Labsphere 98% reflection standard (OOIIrrad software ver. 2.04). The peak reflectance for the yellow paint was around 610 nm and the black paint showed a flat curve. These measurements are very close to those observed in vespid wasps (Kauppinen and Mappes 2003), which are considered to be the model for the mimetic hoverfly, S. longicornis. We detected no ultraviolet reflectance in our paints.
Experimental details
Overall, we ran 8 separate experiments on wild carrot and 9 experiments on goldenrod using both natural and artificial specimens (see Table 1 for details). Each experiment was repeated 12–28 times, reflecting a combined total of approximately 57 h of continual observation.
|
Experiments 1–4: flowers containing a specimen versus unoccupied flowers
In the first set of 4 experiments on goldenrod (experiments 1a–4a) and wild carrot (experiments 1b–4b), one of the 2 flowers was pinned with a specimen, whereas the other one was left unoccupied. Thus, in experiment 1(a, b) one of the flowers contained a pinned wasp, Dolichovespula arenaria; in experiment 2(a, b) one of the flowers had a pinned wasp-mimetic hoverfly, S. longicornis, and in experiment 3(a, b) one of the flowers held a pinned nonmimetic fly, Sarcophaga spp.—no specimen was pinned on the matching flower in each case. In experiment 4(a, b) we placed a black-painted artificial specimen on one flower and left the other one unoccupied.
Experiments 5–9: pinned flowers versus pinned flowers
In our second set of 5 experiments, we compared events on 2 flowers, each harboring a different artificial or natural dried specimen. Experiments 5–8 were again conducted on both goldenrod (5a–8a) and wild carrot (5b–8b) flowers. Experiment 9 was conducted on goldenrod flowers only (the restricted flowering period for carrot prevented us from conducting this final experiment).
In experiment 5(a, b), one of the flowers was pinned with a nonmimetic fly, Sarcophaga spp., and the other flower was pinned with a black-and-yellow wasp specimen, D. arenaria. In experiment 6(a, b), we presented a flower pinned with a nonmimetic fly, Sarcophaga spp., and a flower pinned with mimetic hoverfly, S. longicornis. In experiment 7(a, b), one of the flowers was pinned with a wasp, D. arenaria, and the other was pinned with a wasp-mimic hoverfly, S. longicornis. In experiment 8(a, b), one of the experimental flowers contained a black-painted artificial specimen and the other contained a black-and-yellow–painted artificial specimen. Finally, Experiment 9a (conducted only on goldenrod flowers) involved comparing events on flowers pinned with a black-painted hoverfly, S. longicornis, to events on flowers pinned with a mock-painted hoverfly, S. longicornis.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Weekly surveys revealed that the most common flower visitors present in and around flowers on the experimental field belonged to the order of Hymenoptera, particularly from the family of Apidae (honeybees and bumblebees). Mimetic hoverflies (Syrphidae), nonmimetic flies (e.g., Sarcophaga spp.), and black-and-yellow wasps (e.g., Vespidae) were also among the common species present in our study site (Table 2).
|
Experiments 1–4: flowers containing a specimen versus unoccupied flowers
In both experiments 1a and 1b, significantly, more insects visited unoccupied flowers than flowers pinned with a wasp in the 10-min observation period (Table 1, Figure 1a,b). Likewise, goldenrod flowers pinned with the mimetic hoverfly S. longicornis received significantly fewer insect visitors in the observation period than goldenrod not containing a specimen (experiment 2a, Table 1, Figure 1a). Wild carrot pinned with the hoverfly S. longicornis similarly received overall fewer visitors than wild carrot flowers not containing a specimen, although the difference was of borderline statistical significance in a 2-tailed test (experiment 2b, Table 1, Figure 1b).
|
In experiment 3(a, b), there was no significant difference between the number of visits to unoccupied flowers compared with the number of visits of the flowers pinned with the nonmimetic fly, Sarcophaga spp., on both goldenrod and wild carrot flowers (Table 1) although a strong trend toward visiting unoccupied flowers still existed (Figure 1a,b). Indeed, the number of visits was significantly different between treatments in both cases if a slightly more powerful yet less conservative paired t-test were conducted on root-transformed frequency of visitors (goldenrod: t19 = 2.742, P = 0.013; wild carrot t18 = 2.304, P = 0.033). In experiment 4(a, b), flower visitors showed a highly significant preference for visiting unoccupied flowers compared with the flowers harboring black-painted artificial specimens (Table 1, Figure 1a,b).
Our observations revealed that pollinators tended to frequently hover around and inspect flowers pinned with a specimen without making any physical contact with either the flower or the specimen. We conducted a parallel series of nonparametric analyses (Wilcoxon signed ranks) to compare the number of investigations made by flower visitors. These analyses overall indicated that visitors investigated the pinned flower in the pair significantly more frequently than the unoccupied flower (Table 1, Figure 2a,b).
|
Experiments 5–9: pinned flowers versus pinned flowers
In experiment 5a, there were significantly fewer visits of other insects to goldenrod pinned with the wasp, D. arenaria, compared with flowers pinned with the nonmimetic fly, Sarcophaga spp. (Table 1, Figure 3). A similar trend was observed on wild carrot (experiment 5b), although the difference was of borderline statistical significance (Table 1, Figure 3).
|
Crucially, in experiment 6(a, b), the number of visits to the flowers pinned with a mimetic hoverfly, S. longicornis, was not significantly different from the number of visits to the flowers pinned with a nonmimetic specimen, Sarcophaga spp. (Table 1, Figure 4). Our results from experiment 7, on both goldenrod (7a) and wild carrot (7b), similarly indicated no significant difference between the number of visits to the flowers pinned with a wasp specimen, D. arenaria, and flowers pinned with the mimetic hoverfly S. longicornis (Table 1). Likewise, in experiment 8(a, b), we observed no significant difference between the number of visits to black-painted beads and the number of visits to black-and-yellow–painted beads (Table 1, Figure 4). Finally, the results from experiment 9a, also indicated no difference between the number of visits to flowers pinned with a mock-painted hoverfly, S. longicornis, and the number of visits to flowers pinned with a black-painted hoverfly (Table 1). Parametric t-tests on the root-transformed number of visitors similarly found no evidence of a treatment effect in any of these central comparisons.
|
Our observations indicated that pollinators inspected both occupied flowers at approximately the same frequency without making any physical contact with the food source and regardless of the type and the color of the pinned specimen. Thus, Wilcoxon tests showed no evidence that any pairs of flowers were investigated more often by flower visitors (see Table 1).
Physical attacks
We regularly observed large-sized Vespula maculata, Polistes fuscatus, and even a hoverfly, Spilomyia sayi, physically attacking our pinned specimens and other flower visitors when they were feeding on the inflorescences. On a few occasions, V. maculata and P. fuscatus even tried to remove the pinned nonmimetic fly, Sarcophaga spp., from the flower in an aggressive manner. By contrast, and despite many hours of observation, we did not record any such behavior toward pinned wasps. Incidental touching of the flower visitors to the pinned specimens sometimes occurred, but the overall frequency of physical attacks (typically 2–3 attacks over 20 or so replicates in any given experiment) were too few to allow detailed statistical analysis.
Flower visitors selectivity
We conducted a series of chi-squared tests of association to evaluate whether unoccupied flowers or flowers pinned with a certain specimen were visited relatively more frequently by a certain taxonomic group of flower visitors (Hymenoptera and/or Diptera—the majority of flower visitors). Such tests necessarily involved combining the number of visits by taxonomic groups within a given experiment, assuming that the data were independent. Separate analyses of the results of individual experiments gave no evidence that particular insect order (Hymenoptera or Diptera) visited one of the 2 flower types over the other (Table 3), although expected counts were frequently less than 5 so the results should be treated with some caution (Sokal and Rohlf 1995).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Goldenrod and wild carrot were the most common flowering species present in our field sites during our study period. These 2 plant species were very attractive to insects and were frequently visited by a variety of species from the orders Diptera and Hymenoptera and sometimes Coleoptera, Hemiptera, and Lepidoptera. The results of our weekly surveys confirmed that the species we had used in our pairwise experiments were present at the sites where the study took place, indicating that flower visitors were likely to have had previous encounters with many of them (Table 2). Using the 2 types of natural flower with different colors and shapes also enabled us to look at the consistency of our results (goldenrod inflorescences are more or less cone shaped and larger compared with wild carrot inflorescences, which are arranged on a flat disc-like form). Furthermore, our incidental observations of aggression exhibited by wasps such as V. maculata and P. fuscatus as the experiments progressed gave us good reason to wonder whether hoverflies might benefit in any competitive sense by resembling a wasp.
It has previously been shown that the presence and size of the occupier influences the flower choice of insect pollinators, although the type of the response to these visual cues varies with species. For example, Richter and Tisch (1999) showed that while V. maculifrons, Vespula vidua, and Vespula consobrina avoid flowers pinned with a wasp specimen, V. germanica and P. fuscatus were attracted to the flowers that were occupied by a wasp specimen. Parrish and Fowler (1983) similarly observed that V. maculifrons tended to avoid pinned specimens and artificially painted drawing pins, whereas V. germanica preferred to feed on the resources that are occupied by a greater number of foragers. In recent experiments, it has been shown that bumblebees can be attracted to feed at novel resources through observing conspecifics (Leadbeater and Chitka 2005), although it remains unclear how bees would behave once learning is complete. Indeed, interactions between bumblebees in the field are often regarded as more antagonistic in nature (Stout et al. 1998). In laboratory experiments designed to understand the significance of dark spots on the umbels of D. carota and Artedia squamata (Apiaceae), Eisikowitch (1980) found that naive houseflies (Musca domestica) were attracted to umbels with a dark central flower. From this, he suggested that the spots serve as a "flycatcher" for insects that are in search of mates, thereby promoting their pollination. More recent fieldwork by Lamborn and Ollerton (2000) has cast doubts over this intriguing hypothesis, however, because seed set did not vary in a consistent way between different types of manipulated D. carota.
Our own results support the view that current occupancy may influence flower choice by a community of pollinators. Thus, we have found that insects visited unoccupied goldenrod flowers significantly more frequently than goldenrod flowers occupied by a wasp, mimetic hoverfly, or an artificial specimen (experiments 1a, 2a, and 4a). A very similar pattern was observed for experiments with wild carrot flowers except that only experiments 1b and 4b resulted in a significant difference in the number of visits. It is important to note that even though the results of experiments 3a for goldenrod and 2b and 3b for wild carrot were nonsignificant, they were close to significance and in the same direction. The overall mean number of visitors tended to be less to wild carrot (presumably a result of their smaller size), which may render each of our comparisons less sensitive to a treatment effect in this particular plant species.
Our analysis also clearly indicated that flower visitors investigated and hovered around the occupied flowers more often compared with unoccupied flowers. Three explanations can be offered for investigating the occupied inflorescence by flower visitors and avoiding any physical contact with the pinned specimen, although they are not mutually exclusive. First, flower visitors may perceive the presence of another species as potential danger. Previous studies have also shown that flower visitors avoid visiting potentially dangerous flowers (Dukas 2001; Dukas and Morse 2003). The second possibility is that flower visitors prefer not to share a food resource with another insect, yet need to confirm it is occupied. Third, it is possible that the presence of another visitor on a flower is an indicator for the quality of the resource, and therefore, flower visitors assess the chance of sharing the food source with or even the possibility of winning an aggressive encounter with the occupier. The fact that the number of inspections is significantly higher for pinned specimens compared with unoccupied flowers clearly suggests that such decisions are made at close quarters rather than from a distance.
Intriguingly, our results revealed that flower visitors have a tendency to avoid landing on flowers pinned with a wasp compared with flowers pinned with a nonmimetic fly. By contrast, visitors arrived at flowers pinned with hoverflies at an approximately equal rate compared with similar flowers pinned with nonmimetic flies (experiments 6a, b). Indeed, goldenrod with mock-painted and black-painted hoverflies attracted visitors at similar rates (experiment 9a), and a similar result was observed for black-painted and yellow-painted beads (experiments 8a, b). Collectively, this suggests that although insects may be dissuaded from landing on flowers that contain another individual, and they may show particularly strong avoidance of wasps (quite possibly because of their distinctive shape or antennae length, see Dukas 2001; Kauppinen and Mappes 2003), the yellow-and-black stripes of a hoverfly do not appear to protect an individual from competition in terms of reducing the rates of arrival of potential competitors.
Although nonsignificant results can arise due to insufficient replication, each of our experiments was relatively well replicated and therefore capable of detecting modest differences with reasonable power. To address this issue in more detail, we have followed the guidelines proposed by Colegrave and Ruxton (2003), who argue that confidence intervals (CIs) are a more useful complement to nonsignificant tests than post hoc power tests (see also Colegrave and Ruxton 2005; Johnson 2005). In all the central nonsignificant comparisons to test for the effectiveness of yellow-and-black markings in reducing competition, the (back transformed) mean root difference in the number of visits between the 2 types of flower was always less than 0.2 visits (and generally considerably less), with the 95% CIs (assuming approximate normality) for this difference always having an upper limit of less than 0.7 visits (black-painted beads vs. yellow-and-black–painted beads: goldenrod, back-transformed mean root difference 0.0059 visits [95% CI –0.094 to 0.212]; carrot, back-transformed mean root difference –0.00006 visits [95% CI –0.341 to 0.323]; nonmimetic fly vs. mimetic hoverfly: goldenrod, back-transformed mean root difference 0.0606 [95% CI –0.083 to 0.609]; carrot, back-transformed mean root difference 0.0088 [95% CI –0.115 to 0.277]; black-painted vs. mock-painted hoverfly; goldenrod, back-transformed mean root difference 0.105 [95% CI –0.011 to 0.568]). Of course, like all nonsignificant experiments, we cannot exclude the possibility that black-and-yellow patterns provide some small degree of competitive benefit to hoverflies—indeed, some tantalizing trends are apparent (e.g., the fact that flowers with nonmimetic flies were significantly preferred over flowers with wasps, yet the number of visits to flowers with wasps and hoverflies did not differ significantly, gives particular reason to be cautious). However, after repeated nonsignificant tests and an examination of sample mean effect sizes, we believe that we are safe in concluding that our data in themselves provide no convincing evidence for a competitive benefit from mimicry in hoverflies.
It is possible that alternative designs may have given different answers. For example, some hoverflies possess elongated antennae (e.g., Sphecomyia vittata and Ceriana signifera)—presumably to mimic the long antennae of yellow-jacket wasps—whereas some other hoverfly species, such as S. longicornis and Temnostoma spp., wave their darkened front legs to mimic the presence and movement of the antennae of their potential wasp models (Waldbauer 1970). Moreover, more perfect mimetic hoverflies also wag their wings to mimic wasp's movements (Waldbauer 1988) or they fly in a similar manner as their related models do (Golding and Edmunds 2000; Golding et al. 2001). Behavioral similarities are potentially important in improving the quality of mimicry to the receivers and in our case to potential competitors. However, in using pinned specimens, we limited ourselves to elucidating benefits of morphological mimicry. Further studies would need to be done to test the possibility that behavioral mimicry is necessary before benefits of reduced competition can be realized.
Studies have also shown that some flower visitors like honeybees and bumblebees (Goulson et al. 2000) and even hoverflies (Reader et al. 2005) use short term olfactory cues to mark flowers that they have already visited and possibly to mediate interaction between species competing for flower resources. Although this could potentially affect the results of some experiments, the 10-min observation time allowed us to record frequent flower visits for our experiments. Furthermore, the 2 flower choices in each trial were presented in a pairwise manner, and they were equally exposed to flower visitors including honeybees and bumblebees.
The idea of competitive mimicry is an intriguing one and merits further empirical testing in this and other systems. Our study has revealed very clear evidence that flower choice by pollinators is sensitive to whether flowers are occupied, and it has provided some evidence to suggest that flowers containing wasps are preferentially avoided by pollinators compared with flowers with nonmimetic flies. Nevertheless, we have found no compelling evidence that floral competition is reduced through the adoption of yellow-and-black markings per se, using both artificial prey and natural prey on 2 different flowering species.
|
ACKNOWLEDGEMENTS |
|---|
We thank Jeff Skevington and Jeff Cumming from the Canadian National Collection of Arthropods and Insects for help in identifying the flies and wasps, Paul Catling from Eastern Cereal and Oilseed Research Centre for help in identifying our experimental plants, Imran Khan, our field assistant, and Graham Smith from the National Wildlife Research Centre of Canada for helping us in measuring the color reflectance and providing us with the related equipment. Chris Beatty, Rod Bain, and David Wilkinson kindly read and commented on our manuscript. This work was funded by Canada Foundation for Innovation and Ontario Innovation Trust equipment grants and Natural Sciences and Engineering Research Council grant to T.N.S.
|
FOOTNOTES |
|---|
Arash Rashed is now at the Department of Biological Sciences, University of Cincinnati, Cincinnati, OH 45221, USA.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Bates HW. (1862) Contributions to an insect fauna of the Amazon valley. Lepidoptera: Heliconidae. Trans Linn Soc Lond 23:495–566.
Biesmeijer JC, Richter JAP, Smeets MAJP, Sommeijer MJ. (1999) Niche differentiation in nectar-collecting stingless bees: the influence of morphology, floral choice and interference competition. Ecol Entomol 24:380–388.[CrossRef]
Branquart E and Hemptinne JL. (2000) Selectivity in the exploitation of floral resources by hoverflies (Diptera: Syrphinae). Ecography 23:732–742.[CrossRef][Web of Science]
Colegrave N and Ruxton GD. (2003) Confidence intervals are a more useful complement to nonsignificant tests than are power calculations. Behav Ecol 14:446–450.
Colegrave N and Ruxton GD. (2005) What hypothesis tests are not: a reply to Johnson. Behav Ecol 16:325.
Cott HC. (1940) Adaptive Colouration in Animals. (Meuthen, London).
Dittrich W, Gilbert F, Green P, McGregor P, Grewcock D. (1993) Imperfect mimicry: a pigeon's perspective. Proc R Soc Lond B Biol Sci 251:195–200.
Dukas R. (2001) Effect of perceived danger on flower choice by bees. Ecol Lett 4:327–333.[CrossRef][Web of Science]
Dukas R and Morse DH. (2003) Crab spiders affect flower visitation by bees. Oikos 101:157–163.[CrossRef][Web of Science]
Edmunds M. (1974) Defence in animals: a survey of anti-predator defences. (Longman, Essex (UK)).
Eisikowitch D. (1980) The role of dark flowers in the pollination of certain umbelliferae. J Nat Hist 14:737–742.[CrossRef]
Gamboa GJ, Grudzien TA, Esplie KE, Bura EA. (1996) Kin recognition pheromones in social wasps: combining chemical and behavioral evidence. Anim Behav 51:625–629.[CrossRef]
Gilbert F. (1981) Foraging ecology of hoverflies (Diptera, Syrphidae): morphology of the mouthparts in relation to feeding on nectar and pollen in some common urban species. Ecol Entomol 6:245–262.[CrossRef]
Gilbert F. (2005) The evolution of imperfect mimicry. Insect evolutionary ecology (CABIIn Fellowes MDE, Holloway GJ, Rolff J (Eds.). , Wallingford (UK))231–288.
Golding YC and Edmunds M. (2000) Behavioural mimicry of honeybees (Apis mellifera) by droneflies (Diptera: Syrphidae: Eristalis spp.). Proc R Soc Lond B Biol Sci 267:903–909.[Medline]
Golding YC, Ennos AR, Edmunds M. (2001) Similarity in flight behaviour between the honeybee Apis mellifera (Hymenoptera: Apidae) and its presumed mimic, the dronefly Eristalis tenax (Diptera: Syrphidae). J Exp Biol 204:139–145.[Abstract]
Goulson D, Stout JC, Langley J, Hughes WOH. (2000) Identity and function of scent marks deposited by foraging bumblebees. J Chem Ecol 26:2897–2911.[CrossRef][Web of Science]
Green PT, Gentle L, Peake TM, Scudamore RE, McGregor PK, Gilbert F, Dittrich WH. (1999) Conditioning pigeons to discriminate naturally lit insect specimens. Behav Processes 46:97–102.[Medline]
Holen ØH and Johnstone RA. (2004) The evolution of mimicry under constraints. Am Nat 164:598–613.[CrossRef][Web of Science][Medline]
Howarth B, Edmunds M, Gilbert F. (2004) Does the abundance of hoverfly (Syrphidae) mimics depend on the numbers of their hymenopteran models? Evolution 58:367–375.[CrossRef][Web of Science][Medline]
Johnson DH. (2005) What hypothesis tests are not: a reply to Colegrave & Ruxton. Behav Ecol 16:323–324.
Kauppinen J and Mappes J. (2003) Why are wasps so intimidating: field experiments on hunting dragonflies (Odonata: Aeshna grandis). Anim Behav 66:505–511.[CrossRef]
Kevan PG and Baker HG. (1983) Insects as flower visitors and pollinators. Annu Rev Entomol 28:407–453.[CrossRef][Web of Science]
Kikuchi T. (1962) Studies on the coaction among insects visiting flowers. I. Ecological groups in insects visiting the chrysanthemum flower, Chrysanthemum leucanthemum. Sci Rep Tohoku Univ Fourth Ser (Biol) 28:17–22.
Lamborn E and Ollerton J. (2000) Experimental assessment of the functional morphology of inflorescences of Daucus carota (Apiaceae): testing the ‘fly catcher effect’. Funct Ecol 14:445–454.[CrossRef]
Leadbeater E and Chittka L. (2005) A new mode of information transfer in bumblebees. Curr Biol 15:447–448.[CrossRef][Web of Science][Medline]
Mostler G. (1935) Beobachtungen zur frage der wespenmimikry. Z Morphol Oekol Tiere 29:381–454.[CrossRef]
Nagamitsu T and Inoue T. (1997) Aggressive foraging of social bees as a mechanism of floral resource partiioningn in an Asian tropical rainforest. Oecologia 110:432–439.[CrossRef][Web of Science]
Nickol M. (1994) Volucella zonaria (Diptera: Syrphidae)in Rheinland-Pfalz: distribution, flower visiting, behaviour, color and ecology and remarks on related species. Mitt Pollichia 81:383–405.
Parrish MD and Flower HG. (1983) Contrasting foraging related behaviours in two sympatric wasps (Vespula maculifrons and V. Germanica). Ecol Entomol 8:185–190.
Rashed A, Beatty CD, Forbes MR, Sherratt TN. (2005) Prey selection by dragonflies in relation to prey size and wasp-like colours and patterns. Anim Behav 70:1195–1202.[CrossRef]
Reader Z, MacLeod I, Elliot PT, Robinson OJ, Manica A. (2005) Inter-order interactions between flower-visiting insects: foraging bees avoid flowers previously visited by hoverflies. J Insect Behav 18:51–57.[CrossRef]
Richter MR. (2000) Social wasp (Hymenoptera: Vespidae) foraging behaviour. Annu Rev Entomol 45:121–150.[CrossRef][Web of Science][Medline]
Richter MR and Tisch VL. (1999) Resource choice of social wasps: influence of presence, size, and species of resident wasps. Insectes Soc 46:131–136.[CrossRef]
Ruxton GD, Sherratt TN, Speed MP. (2004) Avoiding attack: the evolutionary ecology of crypsis, warning signals and mimicry. (Oxford University Press, New York).
Schoener TW. (1974) Resource partitioning in ecological communities. Science 185:27–39.
Sokal RR and Rohlf FJ. (1995) Biometry. 3rd ed (W. H. Freeman and Company, New York).
Stout JC, Goulson D, Allen JA. (1998) Repellent scent-marking of flowers by a guild of foraging bumblebees (Bombus spp.). Behav Ecol Sociobiol 43:317–326.[CrossRef][Web of Science]
Waldbauer GP. (1970) Mimicry of hymenopteran antennae by Syrphidae. Psyche 77:45–49.[Medline]
Waldbauer GP. (1988) Insects through the seasons. (Harvard University Press, Cambridge (MA)).
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||