Behavioral Ecology Advance Access originally published online on July 31, 2006
Behavioral Ecology 2006 17(6):947-951; doi:10.1093/beheco/arl028
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Do the multiple defense chemicals of visually distinct species enhance predator learning?
School of Biology, Department of Psychology, Henry Wellcome Building for Neuroecology, University of Newcastle, Framlington Place, Newcastle upon Tyne NE2 4HH, UK
Address correspondence to J. Skelhorn. E-mail: john.skelhorn{at}ncl.ac.uk.
Received 15 December 2005; revised 18 April 2006; accepted 3 July 2006.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Müllerian mimicry, where 2 unpalatable species share a warning pattern, is classically believed to be a form of mutualism, where the species involved share the cost of predator education. Birds learn to avoid a color signal faster when individual prey possesses 1 of 2 bitter-tasting chemicals rather than all having the same chemical, suggesting that Müllerian mimics that possess different defense chemicals are better protected than those that possess the same defense chemical. Using domestic chicks as predators and flavored, colored crumbs for prey, we investigated whether birds learn to avoid 2 visually distinct crumb types faster when each crumb type possesses a different defense chemical than when both crumb types share the same defense chemical. We found that birds learned to avoid 2 visually distinct color signals at a similar rate, irrespective of whether each color signal represented a different defense chemical or whether both color signals represented the same defense chemical. This experiment, therefore, indicates that in terms of predator avoidance learning, possessing 2 defense chemicals is more advantageous when prey look the same than when they look different. This suggests that Müllerian mimics with different defense chemicals not only are better protected than Müllerian mimics that share a single chemical but also benefit more from their mimetic resemblance.
Key words: domestic chick, Müllerian mimicry, novelty, perception, taste.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Aposematic insect species advertise their unpalatability to potential predators using conspicuous coloration (Cott 1940
; Edmunds 1974). Müllerian mimics are sympatric aposematic species that share the same or similar warning patterns (Wickler 1968). These species are classically thought to benefit by sharing the cost of predator education: if a predator learns to avoid a warningly colored species during a fixed number of encounters, then Müllerian mimics benefit as fewer individuals of each species will be killed educating naive predators (Müller 1879).
The evolution of Müllerian mimicry is thought to be a 2-step process. First, a relatively large mutation in one aposematic species causes its bearer to resemble another sympatric aposematic species (henceforth the model species). If the resemblance is sufficiently accurate, predators may generalize their learned avoidance of the model species to the mimetic mutants. If the mutants are better protected than nonmimetic conspecifics, for example, if the model species is more common or more unpalatable than the species containing the mimetic mutants, selection will favor the evolution of Müllerian mimicry (Turner 1995). Because the degree to which mimics are protected from avian predators increases as the accuracy of the resemblance increases (Rowe et al. 2004), mimetic individuals that most accurately resemble the model species will be at a selective advantage, and thus, the resemblance will gradually increase in accuracy (Turner 1987).
Recent mathematical models have suggested that the exact predictions of classical mimicry theory are sensitive to small changes in the learning and forgetting algorithms used to simulate predator behavior (Speed 1993, 1999; Speed and Turner 1999). However, the models do not consider how the interaction of "different" defense chemicals may influence predator learning and memory. This may be an important oversight because some of the best-studied Müllerian mimics (e.g., the Monarch and the Viceroy butterflies) appear to differ in their defense chemistry (Blum 1981; Brower 1984; Nishida 2002).
In a recent series of experiments with domestic chicks (Gallus gallus domesticus), we found that birds learned to avoid a color signal faster when individual prey possesses 1 of 2 bitter-tasting chemicals rather than when all individuals have the same chemical (Skelhorn and Rowe 2005a). This suggests that Müllerian mimics that possess different defense chemicals are better protected than those that possess the same defense chemical (Skelhorn and Rowe 2005a). One explanation for this is that taste novelty increases the attention paid to a single visual signal (Skelhorn and Rowe 2005a). If this is the case, we might also expect the presence of 2 different defense chemicals to enhance avoidance learning of 2 different visual signals, meaning that 2 visually distinct aposematic species that possess different defense chemicals may also be better protected than species that possess the same defense chemical.
We need to understand whether defense chemicals interact to influence the survival of visually distinct aposematic species in order to understand the benefit of mimetic resemblances. This is because the benefit of a mimetic resemblance is calculated by comparing the difference in protection afforded to an individual with the original visual signal and the protection afforded by the mimetic signal. However, no study has yet compared how avian predators learn to avoid visually distinct aposematic prey with either the same or different defense chemicals.
In the following experiment, we used a laboratory system of domestic chicks (G. g. domesticus) foraging on colored chick crumbs, to test the prediction that birds learn to avoid 2 visually distinct aposematic species faster when each species possesses a different defense chemical than when both species share the same defense chemical.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Subjects and housing
Fifty domestic chicks (G. g. domesticus) of mixed sex were hatched in the laboratory and housed in 2 cages measuring 100 x 50 x 50 cm. One cage housed the 36 experimental chicks and the other the 14 "buddy" chicks (see below). They were all subject to a 14:10 h light:dark cycle using uncovered florescent lights with no UV component, and temperatures were maintained at 24–25 °C using room heaters and heat lamps. Water was provided ad lib, as were brown chick starter crumbs, except during training and experimenting when food deprivation was necessary. All deprivation periods were in accordance with Home Office regulations and guidelines. At the end of the experiment, all chicks were donated to free-range smallholdings.
Artificial prey
We produced palatable crumbs and 2 types of bitter-tasting defended crumbs by spraying 150 g of brown starter crumbs with 100 ml of water, 2% quinine sulfate solution, or one drop of 2.5% Bitrex solution made up to 125 ml with water. These concentrations of quinine and Bitrex were chosen as they are known to be equally aversive to domestic chicks (Skelhorn and Rowe 2005a). The crumbs were then allowed to dry before green, red, and black palatable crumbs were produced by spraying 150 g of palatable crumbs with 0.5 ml of sugarflair spruce-green food dye diluted to 90 ml with water, 2 ml of supercook red food dye diluted to 90 ml with water, or 8 ml of supercook black food dye diluted to 90 ml with water. We also produced 2 different types of defended red crumbs by spraying 150 g of quinine-flavored crumbs and 150 g of Bitrex-flavored crumbs with 90 ml of the same solution of red dye and 2 different types of defended black crumbs by spraying 150 g of each defended crumb type with 90 ml of the black dye solution. Crumbs were then allowed to dry before being sieved to ensure they were of a similar size.
Experimental Arena
A cage identical to those used to house the chicks was used as an experimental arena. It was divided into 2 sections separated by a wire mesh screen placed 25 cm from one end of the cage. The main section was used as the experimental arena, the floor of which was green laminated cardboard divided into 80 equally sized rectangles that enabled individually placed crumbs to be identified by their position. The smaller section was used for 2 buddy chicks, which had free access to food and water during trials, and provided social contact for experimental chicks (these chicks were changed every 3 trials). On the first and second days posthatch, the experimental chicks were trained to eat brown chick crumbs from the green floor of the experimental arena (for further details, see Skelhorn and Rowe 2005a).
Testing
On day 3, experimental chicks were assigned to 1 of 4 groups, ensuring that chicks of each sex were distributed equally across groups. After approximately 1.5 h of food deprivation, chicks were individually placed in the experimental arena where they encountered 20 palatable green crumbs, 10 unpalatable red crumbs, and 10 unpalatable black crumbs. Chicks in each of the 4 groups differed in the type of unpalatable crumbs that they received: chicks in the "rQbQ" group received 10 red quinine-flavored crumbs and 10 black quinine-flavored crumbs, chicks in the "rBbB" group received 10 red Bitrex-flavored crumbs and 10 black Bitrex-flavored crumbs, chicks in the "rQbB" group received 10 red quinine-flavored crumbs and 10 black Bitrex-flavored crumbs, and chicks in the "rBbQ" group received 10 red Bitrex-flavored crumbs and 10 black quinine-flavored crumbs.
Crumbs were placed singly in the rectangles drawn on the floor of the experimental arena. The position of each crumb was determined by randomly generated maps produced prior to the experiment. Chicks were allowed to attack (peck or eat) 16 crumbs before being removed from the arena. All chicks received 7 trials in total: 2 each on days 3, 4, and 5 and 1 on day 6.
On day 10, 96 h after the final learning trial, chicks received a single Extinction Trial (Trial 8) to test how chicks recalled their learned aversions. They were presented with 20 palatable green crumbs, 10 palatable red crumbs, and 10 palatable black crumbs in the arena, and the order and color of the crumbs that each chick attacked was recorded.
Statistical analysis
All the measures of learning and memory were analyzed using the same general linear model in which the number of defense chemicals, whether or not the red crumbs were quinine flavored, and the interaction of these 2 factors were included as fixed factors. Using this model, we were able to test the prediction that birds learned to avoid 2 visually distinct unpalatable crumb types faster when each crumb type possessed a different defense chemical than when both crumb types shared the same defense chemical.
Power analysis
We conducted a priori power analyses to determine the sample size that would be necessary to find significant differences in our measures of learning and memory among the experimental groups, if the effect sizes in this experiment were similar to those measured by Skelhorn and Rowe (2005a). Skelhorn and Rowe (2005a) found differences among the experimental groups in the total number of unpalatable crumbs attacked, the difference in the number of unpalatable crumbs attacked between the final learning trial and the Extinction Trial, the difference in the number of crumbs attacked before the first unpalatable crumb between the final learning trial and the Extinction Trial, and the proportion of unpalatable crumbs attacked that were eaten (Skelhorn 2005): the size of these effects were 0.8, 0.7, 0.7, and 0.6, respectively (following Cohen 1988). We then conducted power analyses for determining the sample sizes necessary when comparing groups, using analysis of variance. We set the required power at 0.8, the significance level at 0.05, the number of groups to be compared at 4, and the effect size at 0.6 because this was the smallest effect size found by Skelhorn and Rowe (2005a). We found that using 9 chicks, we could be at least 80% confident that if groups differed significantly, and the effect sizes in this experiment were similar to those seen in Skelhorn and Rowe (2005a), we would find significant differences among our experimental groups.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Learned avoidance of unpalatable prey
All groups acquired the discrimination and learned to avoid both the red and the black defended crumbs by the end of the learning phase (see Figure 1). To investigate the effect of chemical and color differences on predator learning rates, we compared the total number of defended crumbs attacked by each chick from the 4 experimental groups. There were no significant effects of the number of defense chemicals, whether or not red crumbs were quinine flavored, or the interaction of these 2 factors on the total number of defended crumbs attacked (F1,34 = 0.01, P = 0.96; F1,34 = 0.66, P = 0.42; F1,34 = 0.06, P = 0.81; see Figure 2).
|
We used the total number of defended crumbs attacked in Trial 7 as our measures of the asymptotic attack levels. These data could not be normalized by transformation, but nonparametric tests showed that there were no significant differences among the groups in the total number of defended crumbs attacked in Trial 7 (
2 = 1.59, df = 3, P = 0.66; see Figure 1). Looking at where the defended crumb fell in the sequence of crumbs attacked in the final learning trial can also be used as a measure of asymptotic performance because it indicates how reluctant chicks were to attack defended crumbs within a single trial. We calculated the number of crumbs attacked before the first defended crumb, assigning a nominal value of 16 to all chicks that only attacked green crumbs. Again, these data could not be normalized by transformation, but nonparametric tests showed that groups did not differ in the number of crumbs attacked before the first defended crumb (2 = 1.31, df = 3, P = 0.73).
Retention of learned avoidance
In order to determine whether chicks had retained their learned aversions for red and black prey, we compared the total number of conspicuous prey (sum of red and black prey) attacked in Trial 7 (the final learning trial) and Trial 8 (the Extinction Trial). We found that chicks attacked significantly more conspicuous prey in Trial 8 than in Trial 7 (Wilcoxon tests; rBbB group, Z = 2.68, n = 9, P = 0.01; rQbQ group, Z = 2.53, n = 9, P = 0.01; rQbB group, Z = 2.38, n = 9, P = 0.02; rBbQ group, Z = 2.50, n = 9, P = 0.01), indicating that chicks had begun to forget their learned aversion for conspicuous prey.
We then calculated the difference in the number of conspicuous crumbs attacked between Trial 7 and Trial 8, as a measure of how well chicks retained their learned aversions for defended prey. There was no significant effect of the number of defense chemicals, whether or not red crumbs were quinine flavored, or the interaction of these 2 factors on the difference in the number of conspicuous crumbs attacked (F1,34 = 1.14, P = 0.29; F1,34 = 0.29, P = 0.60; F1,34 = 0.0, P = 1.0; see Figure 3), suggesting that there were no differences in how the groups retained their learned aversions.
|
In Trial 8, both the red and the black crumbs were palatable; therefore, the number of red and black crumbs attacked in the Extinction Trial may have been dependent on both memory and relearning about the palatability of the crumbs. To remove any effect of relearning, we calculated the difference in the number of crumbs attacked before the first conspicuous crumb attacked between Trial 7 and Trial 8, as a measure of how well chicks retained their learned aversions for conspicuous prey. There was no significant effect of the number of defense chemicals, whether or not red crumbs were quinine flavored, or the interaction of these 2 factors on the difference in the number of crumbs attacked before the first conspicuous crumb (F1,34 = 0.08, P = 0.78; F1,34 = 0.17, P = 0.68; F1,34 = 0.84, P = 0.37; see Figure 4), suggesting that there was no difference between how the groups retained their learned aversions.
|
Survival of chemical morphs
In order to determine whether birds selectively rejected prey based on their chemical defenses, we compared the proportions of each crumb type that were eaten once they had been attacked. As would be expected, chicks in all groups ate a higher proportion of the palatable crumbs attacked than the defended crumbs attacked (paired t-tests; rBbB group, t = 17.18, df = 8, P < 0.001; rQbQ group, t = 11.16, df = 8, P < 0.001; rQbB group, t = 17.26, df = 8, P < 0.001; rBbQ group, t = 9.55, df = 8, P < 0.001; see Figure 5). However, chicks did not differ in the proportion of defended red crumbs attacked that were eaten and the proportion of defended black crumbs attacked that were eaten (paired t-tests; rBbB group, t = 0.32, df = 8, P = 0.75; rQbQ group, t = 1.14, df = 8, P = 0.29; rQbB group, t = 0.24, df = 8, P = 0.81; rBbQ group, t = 1.22, df = 8, P = 0.26), indicating that chicks found red and black crumbs, and quinine and Bitrex crumbs, equally defended. In addition, there was no significant effect of the number of defense chemicals, whether or not red crumbs were quinine flavored, or the interaction of these 2 factors on the proportion of defended crumbs eaten after attack (F1,34 = 1.17, P = 0.29; F1,34 = 0.001, P = 0.98; F1,34 = 0.86, P = 0.36; see Figure 5). Therefore, having 2 different chemicals instead of one did not reduce the rate at which chicks consumed the defended prey. This result was not confounded by differences in consumption rates between groups because there was no significant effect of the number of defense chemicals, whether or not red crumbs were quinine flavored, or the interaction of these 2 factors on the proportion of green crumbs eaten after attack (F1,34 = 0.69, P = 0.41; F1,34 = 0.29, P = 0.59; F1,34 = 2.57, P = 0.12).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Skelhorn and Rowe (2005a)
demonstrated that chicks learned to avoid a single-color signal faster, and retained the aversion better, when individual prey could possess 1 of 2 bitter-tasting unpalatable chemicals rather than when they all had the same chemical. However, we found no detectable differences in the speed at which birds learned to avoid 2 visually distinct color signals, the retention of that aversion, or the consumption of defended individuals, irrespective of whether each chemically defended color signaled the same or different defense chemicals. This suggests that the chicks learned to avoid the defended crumbs on the basis of their visual signals at similar rates. Alternatively, differences between groups may have been small, and the sample size may not have been large enough to detect them. However, power analyses revealed that the sample size was large enough to detect differences among the experimental groups if the differences were of the same size as those found by Skelhorn and Rowe (2005a). This experiment, therefore, indicates that in terms of predator avoidance learning, possessing 2 defense chemicals is more advantageous when prey look the same than when they look different. It is unclear why possessing 2 different defense chemicals should enhance learning and memory more when species are visually identical Müllerian mimics than when they are visually distinct aposematic species. One potential explanation is that if birds are prepared to ingest a certain "safe" dosage of a particular toxin, birds may find prey with unpredictable defenses more aversive because unpredictability reduces birds' abilities to regulate their toxin intake. When visually identical Müllerian mimics can possess 1 of 2 different defense chemicals, birds cannot predict the consequences of attacking an individual with the shared visual signal. However, when visually distinct species possess different defense chemicals, the outcome of attacking each species remains predictable. Birds may therefore perceive Müllerian mimics with different defense chemicals as more aversive than visually distinct species with different defense chemicals because their defenses are more unpredictable. As a result, birds may learn to avoid Müllerian mimics with different defense chemicals more quickly than visually distinct species with different defense chemicals. However, this hypothesis implies that learning should be dependent on chemical intake. If this were the case, we would expect birds given 1 defense chemical to learn to avoid conspicuous crumbs more quickly than birds given 2 different chemicals because they are ingesting twice as much of the chemical. Because we found no such difference, it seems unlikely that this hypothesis can explain our findings, and it remains unclear why possessing 2 different defense chemicals should enhance learning and memory more when prey types are visually identical than when they are visually distinct.
Irrespective of the exact mechanism, our findings indicate that the defense chemicals possessed by aposematic species may influence the benefit of the Müllerian resemblance. The benefit of a mimetic resemblance is calculated by comparing the protection afforded to an individual by the original distinct visual signal with the protection afforded by the mimetic signal. If we consider a situation where Müllerian mimicry evolves in 2 visually distinct aposematic species, then the results from this experiment demonstrate that the protection afforded by the original visual signals is approximately equal when species have different defense chemicals and when they share a defense chemical. Therefore, there are no additional benefits in terms of learned aversions of distinct prey having either the same or different defense chemistry. However, our previous experiment demonstrated that Müllerian mimics that shared a single defense chemical were approximately twice as likely to be attacked as those with different defense chemicals. This indicates that there should be more benefits to evolving shared coloration when prey have different chemicals than when they share the same defense chemicals. This is because there will be a relatively higher increase in the protection of mimics with different chemicals compared with mimics that have the same ones.
Recent models of Müllerian mimicry considering variation in palatability along a single chemical dimension have generated the prediction that less defended mimics may raise predation rates on their more defended models (Speed 1993, 1999; Speed and Turner 1999). This experiment suggests that the predictions of such models may be altered significantly if Müllerian mimics differ in the defense chemicals they possess because these models are likely to have underestimated the benefit of Müllerian mimicry when mimics possess different defense chemicals. It is, however, still unclear to what degree chemical differences may enhance the benefit of Müllerian mimicry and, as a result, whether the benefit of having different defense chemicals could overcome the costs associated with having a less defended mimic.
Although our data clearly demonstrate that prey with different defense chemicals ultimately benefit more from mimicry than prey that share the same defense chemical, it is unclear whether possessing 2 different defense chemicals actually facilitates the initial evolution of Müllerian mimicry. This is because initially rare mutants acquiring the mimetic morph may not make a significant contribution to the chemical composition of the mimicry complex. As a result, mimetic mutants would only be sampled rarely, and learning may not be enhanced. However, when individuals with different chemicals are visually identical, the probability of surviving a predatory attack is determined by the frequency of the chemical morph in the population of aposematic prey: with individuals with rarer defense chemicals being more likely to survive predatory attacks than individuals with common defense chemicals (Skelhorn and Rowe 2005b). Therefore, although rare mimetic mutants may not benefit from enhanced predator learning, they may benefit from an increased probability of surviving a predatory attack. However, this relies on the assumption that predators use the relative frequencies of different chemicals in an aposematic population, and not in the environment as a whole, to make their decisions about whether to reject prey on attack. This, therefore, needs to be tested experimentally.
Little is known about how the visual and chemical signals of aposematic insects interact to influence predator learning and memory, but the results of this experiment suggest that these interactions may be complex. Taken together with the data of Skelhorn and Rowe (2005a), these data suggest that insects' visual signals may interact with their chemical signals to influence how birds perceive, learn, and forget about variations in chemical defense. They also highlight the importance of considering how the interaction of defense chemicals may have influenced the evolution of Müllerian mimicry.
|
|
ACKNOWLEDGEMENTS |
|---|
We would like to thank Lin Hedgecock and Michelle Waddle who looked after our animals. J.S. is supported by a School of Biology studentship, and C.R. holds a Royal Society Dorothy Hodgkin Research Fellowship. This work was supported by the Royal Society, the Biotechnology and Biological Sciences Research Council and a Wellcome Trust Joint Infrastructure Fund Award.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Blum MS. (1981) Chemical defences of arthropods(Academic Press, New York).
Brower LP. (1984) Chemical defences in Butterflies. In Vane-Write RI and Ackery PR (Eds.). The biology of butterflies(Academic, London) pp. 109–34.
Cohen J. (1988) Statistical power analyses for the behavioral sciences 2nd ed (Academic Press, New York).
Cott HB. (1940) Adaptive colouration in animals(Methuen, London).
Edmunds M. (1974) Defence in animals(Longman, Harlow, UK).
Müller F. (1879) Ithuna and Thyridia: a remarkable case of mimicry in butterflies. Trans R Entomol Soc Lond 1879:xx–xxix.
Nishida R. (2002) Sequestration of defensive substances from plants by Lepidoptera. Annu Rev Entomol 47:57–92.[CrossRef][ISI][Medline]
Rowe C, Lindström L, Lyytinen A. (2004) The importance of pattern similarity between Müllerian mimics in predator avoidance learning. Proc R Soc Lond B Biol Sci 271:407–13.[Medline]
Skelhorn J. (2005) Avian chemical perception and the evolution of insect defences [dissertation](University of Newcastle, Newcastle upon Tyne, UK).
Skelhorn J and Rowe C. (2005a) Tasting the difference: do multiple defence chemicals interact in Müllerian mimicry? . Proc R Soc Lond B Biol Sci 272:339–45.[Medline]
Skelhorn J and Rowe C. (2005b) Frequency-dependent taste-rejection by avian predators may select for defence chemical polymorphisms in aposematic prey. Biol Lett 1:500–3.
Speed MP. (1993) Müllerian mimicry and the psychology of predation. Anim Behav 45:571–80.[CrossRef]
Speed MP. (1999) Robot predators in virtual ecologies: the importance of memory in mimicry studies. Anim Behav 57:203–13.[CrossRef][ISI][Medline]
Speed MP and Turner JRG. (1999) Learning and memory in mimicry: II. Do we understand the mimicry spectrum? Biol J Linn Soc Lond 67:281–312.[CrossRef]
Turner JRG. (1987) The evolutionary dynamics of Batesian and Mullerian mimicry: similarities and differences. Ecol Entomol 12:81–95.
Turner JRG. (1995) Mimicry as a model for coevolution. In Arai R, Kato M, Doi Y (Eds.). Biodiversity and evolution. Contributions to the International Biology Symposium of 1994(National Science Museum Foundation, Tokyo, Japan) pp. 131–50.
Wickler W. (1968) Mimicry in plants and animals(Weidenfeld & Nicolson, London).
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. G. Halpin, J. Skelhorn, and C. Rowe Being conspicuous and defended: selective benefits for the individual Behav. Ecol., September 1, 2008; 19(5): 1012 - 1017. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||