The effect of alternative prey on the dynamics of imperfect Batesian and Müllerian mimicries

Both Batesian and Müllerian mimicries are considered classical evidence of natural selection where predation pressure has, at times, created a striking similarity between unrelated prey species. Batesian mimicry, in which palatable mimics resemble unpalatable aposematic species, is parasitic and only beneficial to the mimics. By contrast, in classical Müllerian mimicry the cost of predators' avoidance learning is shared between similar unpalatable co-mimics, and therefore mimicry benefits all parties. Recent studies using mathematical modeling have questioned the dynamics of Müllerian mimicry, suggesting that fitness benefits should be calculated in a way similar to Batesian mimicry; that is, according to the relative unpalatability difference between co-mimics. Batesian mimicry is very sensitive to the availability of alternative prey, but the effects of alternative prey for Müllerian dynamics are not known and experiments are rare. We designed two experiments to test the effect of alternative prey on imperfect Batesian and Müllerian mimicry complexes. When alternative prey were scarce, imperfect Batesian mimics were selected out from the population, but abundantly available alternative prey relaxed selection against imperfect mimics. Birds learned to avoid both Müllerian models and mimics irrespective of the availability of alternative prey. However, the rate of avoidance learning of models increased when alternative prey were abundant. This experiment suggests that the availability of alternative prey affects the dynamics of both Müllerian and Batesian mimicry, but in different ways.     

Alternative prey can change model–mimic dynamics between parasitism and mutualism

Classical (conventional) Müllerian mimicry theory predicts that two (or more) defended prey sharing the same signal always benefit each other despite the fact that one species can be more toxic than the other. The quasi‐Batesian (unconventional) mimicry theory, instead, predicts that the less defended partner of the mimetic relationship may act as a parasite of the signal, causing a fitness loss to the model. Here we clarify the conditions for parasitic or mutualistic relationships between aposematic prey, and build a model to examine the hypothesis that the availability of alternative prey is crucial to Müllerian and quasi‐Batesian mimicry. Our model is based on optimal behaviour of the predator. We ask if and when it is in the interest of the predator to learn to avoid certain species as prey when there is alternative (cryptic) prey available. Our model clearly shows that the role of alternative prey must be taken into consideration when studying model–mimic dynamics. When food is scarce it pays for the predator to test the models and mimics, whereas if food is abundant predators should leave the mimics and models untouched even if the mimics are quite edible. Dynamics of the mimicry tend to be classically Müllerian if mimics are well defended, while quasi‐Batesian dynamics are more likely when they are relatively edible. However, there is significant overlap: in extreme cases mimics can be harmful to models (a quasi‐Batesian case) even if the species are equally toxic. A crucial parameter explaining this overlap is the search efficiency with which indiscriminating vs. discriminating predators find cryptic prey. Quasi‐Batesian mimicry becomes much more likely if discrimination increases the efficiency with which the specialized predator finds cryptic prey, while the opposite case tends to predict Müllerian mimicry. Our model shows that both mutualistic and parasitic relationship between model and mimic are possible and the availability of alternative prey can easily alter this relationship.     

The importance of pattern similarity between Müllerian mimics in predator avoidance learning

Müllerian mimicry, where unpalatable prey share common warning patterns, has long fascinated evolutionary biologists. It is commonly assumed that Müllerian mimics benefit by sharing the costs of predator education, thus reducing per capita mortality, although there has been no direct test of this assumption. Here, we specifically measure the selection pressure exerted by avian predators on unpalatable prey with different degrees of visual similarity in their warning patterns. Using wild-caught birds foraging on novel patterned prey in the laboratory, we unexpectedly found that pattern similarity did not increase the speed of avoidance learning, and even dissimilar mimics shared the education of naive predators. This was a consistent finding across two different densities of unpalatable prey, although mortalities were lower at the higher density as expected. Interestingly, the mortalities of Müllerian mimics were affected by pattern similarity in the predicted way by the end of our experiment, although the result was not quite significant. This suggests that the benefits to Müllerian mimics may emerge only later in the learning process, and that predator experience of the patterns may affect the degree to which pattern similarity is important. This highlights the need to measure the behaviour of real predators if we are to understand fully the evolution of mimicry systems.     

Tasting the difference: do multiple defence chemicals interact in Müllerian mimicry?

Müllerian mimicry, where two 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. The evolutionary dynamics of Müllerian mimicry have recently become a controversial subject, after mathematical models have shown that if minor alterations are made to assumptions about the way in which predators learn and forget about unpalatable prey, this textbook case of mutualism may not be mutualistic at all. An underlying assumption of these models is that Müllerian mimics possess the same defence chemical. However, some Müllerian mimics are known to possess different defence chemicals. Using domestic chicks as predators and coloured crumbs flavoured with either the same or different unpalatable chemicals as prey, we provide evidence that two defence chemicals can interact to enhance predator learning and memory. This indicates that Müllerian mimics that possess different defence chemicals are better protected than those that share a single defence chemical. These data provide insight into how multiple defence chemicals are perceived by birds,and how they influence the way birds learn and remember warningly coloured prey. They highlight the importance of considering how different toxins in mimicry rings can interact in the evolution and maintenance of Müllerian mimicry and could help to explain the remarkable variation in chemical defences found within and between species.     

Causes and Consequences of a Lack of Coevolution in Müllerian mimicry

Müllerian mimicry, in which both partners are unpalatable to predators, is often used as an example of a coevolved mutualism. However, it is theoretically possible that some Müllerian mimics are parasitic if a weakly defended mimic benefits at the expense of a more highly defended model, a phenomenon known as ‘quasi-Batesian mimicry’. The theory expounded by Müller and extended here for unequal unpalatability, on the other hand, suggests that quasi-Batesian mimicry should be rare in comparison with classical, or mutualistic Müllerian mimicry. Evolutionarily, quasi-Batesian mimicry has consequences similar to classical Batesian mimicry, including unilateral ‘advergence’ of the mimic to the model, and diversifying frequency-dependent selection on the mimic which may lead to mimetic polymorphism. In this paper, theory and empirical evidence for mutual benefit and coevolution in Müllerian mimicry are reviewed. I use examples from well-known insect Müllerian mimicry complexes: the Limenitis–Danaus (Nymphalidae) system in North America, the Bombus–Psithyrus (Apidae) system in the north temperate zone, and the Heliconius–Laparus (Nymphalidae) system in tropical America. These give abundant evidence for unilateral advergence, and no convincing evidence, to my knowledge, for coevolved mutual convergence. Furthermore, mimetic polymorphisms are not uncommon. Yet classical mutualistic Müllerian mimicry, coupled with spatial (and possibly temporal) variation in model abundances convincingly explain these apparent anomalies without recourse to a quasi-Batesian explanation. Nevertheless, the case against classical Müllerian mimicry is not totally disproved, and should be investigated further. I hope that this tentative analysis of actual mimicry rings may encourage others to look for evidence of coevolution and quasi-Batesian effects in a variety of other Müllerian mimicry systems. This revised version was published online in July 2006 with corrections to the Cover Date.     

Natural selection in mimicry

Biological mimicry has served as a salient example of natural selection for over a century, providing us with a dazzling array of very different examples across many unrelated taxa. We provide a conceptual framework that brings together apparently disparate examples of mimicry in a single model for the purpose of comparing how natural selection affects models, mimics and signal receivers across different interactions. We first analyse how model–mimic resemblance likely affects the fitness of models, mimics and receivers across diverse examples. These include classic Batesian and Müllerian butterfly systems, nectarless orchids that mimic Hymenoptera or nectar‐producing plants, caterpillars that mimic inert objects unlikely to be perceived as food, plants that mimic abiotic objects like carrion or dung and aggressive mimicry where predators mimic food items of their own prey. From this, we construct a conceptual framework of the selective forces that form the basis of all mimetic interactions. These interactions between models, mimics and receivers may follow four possible evolutionary pathways in terms of the direction of selection resulting from model–mimic resemblance. Two of these pathways correspond to the selective pressures associated with what is widely regarded as Batesian and Müllerian mimicry. The other two pathways suggest mimetic interactions underpinned by distinct selective pressures that have largely remained unrecognized. Each pathway is characterized by theoretical differences in how model–mimic resemblance influences the direction of selection acting on mimics, models and signal receivers, and the potential for consequent (co)evolutionary relationships between these three protagonists. The final part of this review describes how selective forces generated through model–mimic resemblance can be opposed by the basic ecology of interacting organisms and how those forces may affect the symmetry, strength and likelihood of (co)evolution between the three protagonists within the confines of the four broad evolutionary possibilities. We provide a clear and pragmatic visualization of selection pressures that portrays how different mimicry types may evolve. This conceptual framework provides clarity on how different selective forces acting on mimics, models and receivers are likely to interact and ultimately shape the evolutionary pathways taken by mimetic interactions, as well as the constraints inherent within these interactions.     

Natural selection on unpalatable species imposed by state-dependent foraging behaviour

Müllerian mimicry is typically thought to arise as a consequence of defended prey species adopting a similar way of signalling their unprofitability, thereby reducing the costs of predator education. Here we consider subsequent selection on the morphology of prey species, in the potentially lengthy period of time when predators are generally aware of the noxious qualities of their prey (and so no further learning is involved). Using a pair of stochastic dynamic programming equations which describe both the toxin burdens of a predator and its energy level, we identified the optimal state-dependent rules that maximize a predator's long-term survivorship, and examined the implications of this behaviour for the evolution of prey morphologies. When palatable prey are in short supply then those prey species which contain relatively low doses of toxins become profitable to consume by hungry predators. Under these conditions, a weakly defended prey could gain selective advantage in the post educational period by resembling a prey species which contained a higher dose of the same or different toxins, although the precise nature of the ecological relationship between model and mimic could either be mutualistic or parasitic depending on how mimic density increases when favoured by selection. Our work formally demonstrates that one does not always need to invoke educational effects to explain why two or more unpalatable species have evolved a similar appearance, or to explain why mimetic similarity among distasteful species is maintained over time. When two species contain high levels of different toxins then they may gain mutual advantage by resembling one another, not only by educating the predator as to their common unprofitability (classical Müllerian mimicry), but also by increasing predator uncertainty as to the specific kind of toxin a prey item contains.     

Prey community structure affects how predators select for Mullerian mimicry

Müllerian mimicry describes the close resemblance between aposematic prey species; it is thought to be beneficial because sharing a warning signal decreases the mortality caused by sampling by inexperienced predators learning to avoid the signal. It has been hypothesized that selection for mimicry is strongest in multi-species prey communities where predators are more prone to misidentify the prey than in simple communities. In this study, wild great tits (Parus major) foraged from either simple (few prey appearances) or complex (several prey appearances) artificial prey communities where a specific model prey was always present. Owing to slower learning, the model did suffer higher mortality in complex communities when the birds were inexperienced. However, in a subsequent generalization test to potential mimics of the model prey (a continuum of signal accuracy), only birds that had foraged from simple communities selected against inaccurate mimics. Therefore, accurate mimicry is more likely to evolve in simple communities even though predator avoidance learning is slower in complex communities. For mimicry to evolve, prey species must have a common predator; the effective community consists of the predator's diet. In diverse environments, the limited diets of specialist predators could create 'simple community pockets' where accurate mimicry is selected for.     

Batesian mimics influence mimicry ring evolution

Mathematical models of mimicry typically involve artificial prey species with fixed colorations or appearances; this enables a comparison of predation rates to demonstrate the level of protection a mimic might be afforded. Fruitful theoretical results have been produced using this method, but it is also useful to examine the possible evolutionary consequences of mimicry. To that end, we present individual-based evolutionary simulation models where prey colorations are free to evolve. We use the models to examine the effect of Batesian mimics on Müllerian mimics and mimicry rings. Results show that Batesian mimics can potentially incite Müllerian mimicry relationships and encourage mimicry ring convergence.     

Müllerian mimicry among bees and wasps: a review of current knowledge and future avenues of research

Many bees and stinging wasps, or aculeates, exhibit striking colour patterns or conspicuous coloration, such as black and yellow stripes. Such coloration is often interpreted as an aposematic signal advertising aculeate defences: the venomous sting. Aposematism can lead to Müllerian mimicry, the convergence of signals among different species unpalatable to predators. Müllerian mimicry has been extensively studied, notably on Neotropical butterflies and poison frogs. However, although a very high number of aculeate species harbour putative aposematic signals, aculeates are under-represented in mimicry studies. Here, we review the literature on mimicry rings that include bee and stinging wasp species. We report over a hundred described mimicry rings, involving a thousand species that belong to 19 aculeate families. These mimicry rings are found all throughout the world. Most importantly, we identify remaining knowledge gaps and unanswered questions related to the study of Müllerian mimicry in aculeates. Some of these questions are specific to aculeate models, such as the impact of sociality and of sexual dimorphism in defence levels on mimicry dynamics. Our review shows that aculeates may be one of the most diverse groups of organisms engaging in Müllerian mimicry and that the diversity of aculeate Müllerian mimetic interactions is currently under-explored. Thus, aculeates represent a new and major model system to study the evolution of Müllerian mimicry. Finally, aculeates are important pollinators and the global decline of pollinating insects raises considerable concern. In this context, a better understanding of the impact of Müllerian mimicry on aculeate communities may help design strategies for pollinator conservation, thereby providing future directions for evolutionary research.     

Investigating Müllerian mimicry: predator learning and variation in prey defences

Inexperienced predators are assumed to select for similarity of warning signals in aposematic species (Müllerian mimicry) when learning to avoid them. Recent theoretical work predicts that if co-mimic species have unequal defences, predators attack them according to their average unpalatability and mimicry may not be beneficial for the better defended co-mimic. In this study, we tested in a laboratory environment whether a uniform warning signal is superior to a variable one in promoting predator learning, and simultaneously whether co-mimics are preyed upon according to their average unpalatability. There was an interaction of signal variation and unpalatability but inexperienced birds did not select for signal similarity in artificial prey; when the prey was moderately defended a variable signal was even learnt faster than a uniform one. Due to slow avoidance learning, moderately defended prey had higher mortality than highly defended prey (although this was not straightforward), but mixing high and moderate unpalatability did not increase predation compared with high unpalatability. This does not support the view that predators are sensitive to varying unpalatability. The results suggest that inexperienced predators may neither strongly select for accurate Müllerian mimicry nor affect the benefits of mimicry when the co-mimics are unequally defended.     

A hypothesis to explain accuracy of wasp resemblances

Mimicry is one of the oldest concepts in biology, but it still presents many puzzles and continues to be widely debated. Simulation of wasps with a yellow‐black abdominal pattern by other insects (commonly called “wasp mimicry”) is traditionally considered a case of resemblance of unprofitable by profitable prey causing educated predators to avoid models and mimics to the advantage of both (Figure 1a). However, as wasps themselves are predators of insects, wasp mimicry can also be seen as a case of resemblance to one's own potential antagonist. We here propose an additional hypothesis to Batesian and Müllerian mimicry (both typically involving selection by learning vertebrate predators; cf. Table 1) that reflects another possible scenario for the evolution of multifold and in particular very accurate resemblances to wasps: an innate, visual inhibition of aggression among look‐alike wasps, based on their social organization and high abundance. We argue that wasp species resembling each other need not only be Müllerian mutualists and that other insects resembling wasps need not only be Batesian mimics, but an innate ability of wasps to recognize each other during hunting is the driver in the evolution of a distinct kind of masquerade, in which model, mimic, and selecting agent belong to one or several species (Figure  1b). Wasp mimics resemble wasps not (only) to be mistaken by educated predators but rather, or in addition, to escape attack from their wasp models. Within a given ecosystem, there will be selection pressures leading to masquerade driven by wasps and/or to mimicry driven by other predators that have to learn to avoid them. Different pressures by guilds of these two types of selective agents could explain the widely differing fidelity with respect to the models in assemblages of yellow jackets and yellow jacket look‐alikes.     

Müllerian mimicry: an examination of Fisher's theory of gradual evolutionary change

In 1927, Fisher suggested that Müllerian mimicry evolution could be gradual and driven by predator generalization. A competing possibility is the so-called two-step hypothesis, entailing that Müllerian mimicry evolves through major mutational leaps of a less-protected species towards a better-protected, which sets the stage for coevolutionary fine-tuning of mimicry. At present, this hypothesis seems to be more widely accepted than Fisher's suggestion. We conducted individual-based simulations of communities with predators and two prey types to assess the possibility of Fisher's process leading to a common prey appearance. We found that Fisher's process worked for initially relatively similar appearances. Moreover, by introducing a predator spectrum consisting of several predator types with different ranges of generalization, we found that gradual evolution towards mimicry occurred also for large initial differences in prey appearance. We suggest that Fisher's process together with a predator spectrum is a realistic alternative to the two-step hypothesis and, furthermore, it has fewer problems with purifying selection. We also examined the factors influencing gradual evolution towards mimicry and found that not only the relative benefits from mimicry but also the mutational schemes of the prey types matter.     

Spatial mosaic formation through frequency-dependent selection in Müllerian mimicry complexes

Although contemporary models of Müllerian mimicry have considered the movement of interfacial boundaries between two distinct mimetic forms, and even the possibility of polymorphisms in two patch systems, no model has considered how multiple forms of Müllerian mimics might evolve and be maintained over large geographical areas. A spatially explicit individual-based model for the evolution of Müllerian mimicry is presented, in which two unpalatable species are distributed over discrete cells within a regular lattice. Populations in each cell are capable of genetic drift and experience localized dispersal as well as frequency-dependent selection by predators. When each unpalatable prey species was introduced into a random cell and allowed to spread, then mimicry evolved throughout the system in the form of a spatial mosaic of phenotypes, separated by narrow "hybrid zones". The primary mechanism generating phenotypic diversity was the occasional establishment of new mutant forms in unoccupied cells and their subsequent maintenance (and spread) through frequency-dependent selection. The mean number of discrete clusters of the same morph that formed in the lattice was higher the higher the intensity of predation, and higher the lower the dispersal rate of unpalatable prey. Under certain conditions the hybrid zones moved, in a direction dependent on the curvature of their interfacial boundaries. However, the mimetic mosaics were highly stable when the intensity of predation was high and the rate of prey dispersal was low. Overall, this model highlights how a stable mosaic of different mimetic forms can evolve from a range of starting conditions through a combination of chance effects and localized frequency-dependent selection.     

FEATURE THEORY AND THE TWO‐STEP HYPOTHESIS OF MÜLLERIAN MIMICRY EVOLUTION

The two‐step hypothesis of Müllerian mimicry evolution states that mimicry starts with a major mutational leap between adaptive peaks, followed by gradual fine‐tuning. The hypothesis was suggested to solve the problem of apostatic selection producing a valley between adaptive peaks, and appears reasonable for a one‐dimensional phenotype. Extending the hypothesis to the realistic scenario of multidimensional phenotypes controlled by multiple genetic loci can be problematic, because it is unlikely that major mutational leaps occur simultaneously in several traits. Here we consider the implications of predator psychology on the evolutionary process. According to feature theory, single prey traits may be used by predators as features to classify prey into discrete categories. A mutational leap in such a trait could initiate mimicry evolution. We conducted individual‐based evolutionary simulations in which virtual predators both categorize prey according to features and generalize over total appearances. We found that an initial mutational leap toward feature similarity in one dimension facilitates mimicry evolution of multidimensional traits. We suggest that feature‐based predator categorization together with predator generalization over total appearances solves the problem of applying the two‐step hypothesis to complex phenotypes, and provides a basis for a theory of the evolution of mimicry rings.     

Diversity in Müllerian mimicry: The optimal predator sampling strategy explains both local and regional polymorphism in prey

The convergent evolution of warning signals in unpalatable species, known as Müllerian mimicry, has been observed in a wide variety of taxonomic groups. This form of mimicry is generally thought to have arisen as a consequence of local frequency‐dependent selection imposed by sampling predators. However, despite clear evidence for local selection against rare warning signals, there appears an almost embarrassing amount of polymorphism in natural warning colors, both within and among populations. Because the model of predator cognition widely invoked to explain Müllerian mimicry (Müller's “fixed ” model) is highly simplified and has not been empirically supported; here, we explore the dynamical consequences of the optimal strategy for sampling unfamiliar prey. This strategy, based on a classical exploration–exploitation trade‐off, not only allows for a variable number of prey sampled, but also accounts for predator neophobia under some conditions. In contrast to Müller's “fixed ” sampling rule, the optimal sampling strategy is capable of generating a variety of dynamical outcomes, including mimicry but also regional and local polymorphism. Moreover, the heterogeneity of predator behavior across space and time that a more nuanced foraging strategy allows, can even further facilitate the emergence of both local and regional polymorphism in prey warning color.     

Aversive reactions of two invertebrate predators to European red–black insects

Prey species gain protection by imitating signals of unpalatable models in defensive mimicry. Mimics have been traditionally classified as Batesian (palatable mimic resembling an unpalatable model) or Müllerian (unpalatable mimic resembling a similarly unpalatable model). However, recent studies suggest that rather than discrete categories, the phenomenon of mimicry can be better understood as a continuum. The level of unpalatability of defended prey is a key factor in determining the type of mimetic relationship. Herein, we used insects (ladybugs and true bugs) from a putative European “red–black” mimetic complex as experimental models of defended species and crickets as a control prey. We offered the prey to two species of sympatric invertebrate predators (praying mantis and spider) and video recorded the interactions. We tested three alternative hypotheses, namely (i) the three red–black species tested are similarly defended against both predators; (ii) some red–black species are better defended than others against both predator species, and (iii) the effectiveness of the red–black species defenses is predator dependent. Both predators attacked all prey types with a similar frequency. But while all three red–black species similarly elicited aversive behaviors in spiders, the mantises' aversive reactions varied depending on the prey species. Our results provide support to the third hypothesis, suggesting that the same prey species can fall into different parts of the spectrum of palatability–unpalatability depending on the type of predator.     

Specialized avian predators repeatedly attack novel color morphs of Heliconius butterflies

The persistence of Müllerian mimicry and geographically distinct wing patterns, as observed in many Heliconius species (Lepidoptera: Nymphalidae), is difficult to explain from a predator's perspective: predator selection against locally rare patterns must persist despite avoidance learning. Maintaining spatial color-pattern polymorphism requires local pattern avoidance, fine-scale discrimination among similar wing patterns, and repeated attacks on novel color patterns. I tested for these behaviors by presenting 80 adult rufous-tailed jacamars (Galbula ruficauda) with three morphs of Heliconius butterflies, and then presenting the same suite of butterflies to 46 of these jacamars between four and 429 days later. These trials offer the first direct evidence of the selective predator behavior required to maintain aposematic polymorphism: jacamars avoid local aposematic morphs while repeatedly attacking similar but novel morphs over time.     

Mutualistic mimicry enhances species diversification through spatial segregation and extension of the ecological niche space

Species richness varies among clades, yet the drivers of diversification creating this variation remain poorly understood. While abiotic factors likely drive some of the variation in species richness, ecological interactions may also contribute. Here, we examine one class of potential contributors to species richness variation that is particularly poorly understood: mutualistic interactions. We aim to elucidate large‐scale patterns of diversification mediated by mutualistic interactions using a spatially explicit population‐based model. We focus on mutualistic Müllerian mimicry between conspicuous toxic prey species, where convergence in color patterns emerges from predators' learning process. To investigate the effects of Müllerian mimicry on species diversification, we assume that some speciation events stem from shifts in ecological niches, and can also be associated with shift in mimetic color pattern. Through the emergence of spatial mosaics of mimetic color patterns, Müllerian mimicry constrains the geographical distribution of species and allows different species occupying similar ecological niches to exist simultaneously in different regions. Müllerian mimicry and the resulting spatial segregation of mimetic color patterns thus generate more balanced phylogenetic trees and increase overall species diversity. Our study sheds light on complex effects of Müllerian mimicry on ecological, spatial, and phylogenetic diversification.     

Biased predation could promote convergence yet maintain diversity within Müllerian mimicry rings of Oreina leaf beetles

Müllerian mimicry is a classic example of adaptation, yet Müller's original theory does not account for the diversity often observed in mimicry rings. Here, we aimed to assess how well classical Müllerian mimicry can account for the colour polymorphism found in chemically defended Oreina leaf beetles by using field data and laboratory assays of predator behaviour. We also evaluated the hypothesis that thermoregulation can explain diversity between Oreina mimicry rings. We found that frequencies of each colour morph were positively correlated among species, a critical prediction of Müllerian mimicry. Predators learned to associate colour with chemical defences. Learned avoidance of the green morph of one species protected green morphs of another species. Avoidance of blue morphs was completely generalized to green morphs, but surprisingly, avoidance of green morphs was less generalized to blue morphs. This asymmetrical generalization should favour green morphs: indeed, green morphs persist in blue communities, whereas blue morphs are entirely excluded from green communities. We did not find a correlation between elevation and coloration, rejecting thermoregulation as an explanation for diversity between mimicry rings. Biased predation could explain within‐community diversity in warning coloration, providing a solution to a long‐standing puzzle. We propose testable hypotheses for why asymmetric generalization occurs, and how predators maintain the predominance of blue morphs in a community, despite asymmetric generalization.