Model aversiveness and the evolution of imperfect Batesian mimics

There are numerous examples of Batesian mimics that only imperfectly resemble their models. Given that inaccurate mimics are known to be predated more frequently than accurate ones, imperfect mimicry therefore poses something of a conundrum. One putative explanation, the relaxed selection hypothesis, predicts that when the cost of attacking a model is high relative to the benefit of consuming a mimic, selection against imperfect mimics will be relaxed, allowing mimics to be more imperfect for a given level of fitness. However, empirical support for this hypothesis is equivocal. Here, we report an experimental test of the relaxed selection hypothesis, in which human participants were tasked with discriminating between artificial stimuli representing models and mimics. In response to "attacking" a model (i.e., misclassifying it as palatable, or non-aversive) they received either a mild electric shock (high cost) or vibratory feedback (low cost). Consistent with the predictions of this hypothesis, we found that when the cost of attacking a model was high, mimetic phenotype could deviate more from the model (i.e., be more imperfect) for a given level of fitness than when the cost of attacking a model was low. Moreover, when the cost of attacking a model was high, participants showed an increased latency to attack. This finding shows that given sufficient costs, the relaxed selection hypothesis is a plausible explanation for the evolution of imperfect mimicry.

Signal detection models as contextual bandits

Signal detection theory (SDT) has been widely applied to identify the optimal discriminative decisions of receivers under uncertainty. However, the approach assumes that decision-makers immediately adopt the appropriate acceptance threshold, even though the optimal response must often be learned. Here we recast the classical normal-normal (and power-law) signal detection model as a contextual multi-armed bandit (CMAB). Thus, rather than starting with complete information, decision-makers must infer how the magnitude of a continuous cue is related to the probability that a signaller is desirable, while simultaneously seeking to exploit the information they acquire. We explain how various CMAB heuristics resolve the trade-off between better estimating the underlying relationship and exploiting it. Next, we determined how naive human volunteers resolve signal detection problems with a continuous cue. As anticipated, a model of choice (accept/reject) that assumed volunteers immediately adopted the SDT-predicted acceptance threshold did not predict volunteer behaviour well. The Softmax rule for solving CMABs, with choices based on a logistic function of the expected payoffs, best explained the decisions of our volunteers but a simple midpoint algorithm also predicted decisions well under some conditions. CMABs offer principled parametric solutions to solving many classical SDT problems when decision-makers start with incomplete information.

Body size in Batesian mimicry

A variety of traits is available for predators to distinguish unpalatable prey from palatable Batesian mimics. Among them, body size has received little attention as a possible mimetic trait. Size should influence predator behaviour if it shows variation between models and mimics, is detectable by the predator in question, and is not overshadowed by other traits more salient to the predator. Simple predictions within mimetic populations are that perfect mimics receive the lowest predation rate. However, prey body size is typically tightly linked to the nutritional yield and handling time for a successful predator, as well as likely being correlated with a model’s levels of defence. In certain circumstances, these confounding factors might mean that (a) selection pressures on a mimic’s size either side of the model’s phenotype are not symmetrical, (b) the optimal body size for a mimic is not necessarily equal to that of the model, and/or (c) for predators, attacking better mimics of a model’s body size more readily is adaptive. I discuss promising avenues for improving our understanding of body size as a mimetic trait, including the importance of treatments that range in both directions from the model’s size. Further work is required to understand how body size ranks in saliency against other mimetic traits such as pattern. Comparative studies could investigate whether mimics are limited to resembling only models that are already similar in size.

Field evidence for colour mimicry overshadowing morphological mimicry

Imperfect mimicry may be maintained when the various components of an aposematic signal have different salience for predators. Experimental laboratory studies provide robust evidence for this phenomenon. Yet, evidence from natural settings remains scarce. We studied how natural bird predators assess multiple features in a multicomponent aposematic signal in the Neotropical 'clear wing complex' mimicry ring, dominated by glasswing butterflies. We evaluated two components of the aposematic signal, wing colouration and wing morphology, in a predation experiment based on artificial replicas of glasswing butterflies (model) and Polythoridae damselflies (mimics) in their natural habitat. We also studied the extent of the colour aposematic signal in the local insect community. Finally, we inspected the nanostructures responsible for this convergent colour signal, expected to highly differ between these phylogenetically distinct species. Our results provide direct evidence for a stronger salience of wing colouration than wing morphology, as well as stronger selection on imperfect than in perfect colour mimics. Additionally, investigations of how birds perceive wing colouration of the local insect community provides further evidence that a UV-reflective white colouration is being selected as the colour aposematic signal of the mimicry ring. Using electron microscopy, we also suggest that damselflies have convergently evolved the warning colouration through a pre-adaptation. These findings provide a solid complement to previous experimental evidence suggesting a key influence of the cognitive assessment of predators driving the evolution of aposematic signals and mimicry rings.

Towards an ecology of protective coloration

The strategies underlying different forms of protective coloration are well understood but little attention has been paid to the ecological, life-history and behavioural circumstances under which they evolve. While some comparative studies have investigated the ecological correlates of aposematism, and background matching, the latter particularly in mammals, few have examined the ecological correlates of other types of protective coloration. Here, we first outline which types of defensive coloration strategies may be exhibited by the same individual; concluding that many protective coloration mechanisms can be employed simultaneously, particularly in conjunction with background matching. Second, we review the ecological predictions that have been made for each sort of protective coloration mechanism before systematically surveying phylogenetically controlled comparative studies linking ecological and social variables to antipredator defences that involve coloration. We find that some a priori predictions based on small-scale empirical studies and logical arguments are indeed supported by comparative data, especially in relation to how illumination affects both background matching and self-shadow concealment through countershading; how body size is associated with countershading, motion dazzle, flash coloration and aposematism, although only in selected taxa; how immobility may promote background matching in ambush predators; and how mobility may facilitate motion dazzle. Examination of nearly 120 comparative tests reveals that many focus on ecological variables that have little to do with predictions derived from antipredator defence theory, and that broad-scale ecological studies of defence strategies that incorporate phylogenetics are still very much in their infancy. We close by making recommendations for future evolutionary ecological research.

Signal detection: applying analysis methods from psychology to animal behaviour

Conspecific acceptance thresholds (Reeve 1989 Am. Nat. 133, 407-435), which have been widely applied to explain ecological behaviour in animals, proposed how sensory information, prior information and the costs of decisions determine actions. Signal detection theory (Green & Swets 1966 Signal detection theory and psychophysics; SDT), which forms the basis of CAT models, has been widely used in psychological studies to partition the ability to discriminate sensory information from the action made as a result of it. In this article, we will review the application of SDT in interpreting the behaviour of laboratory animals trained in operant conditioning tasks and then consider its potential in ecological studies of animal behaviour in natural environments. Focusing on the nest-mate recognition systems exhibited by social insects, we show how the quantitative application of SDT has the potential to transform acceptance rate data into independent indices of cue sensitivity and decision criterion (also known as the acceptance threshold). However, further tests of the assumptions underlying SDT analysis are required. Overall, we argue that SDT, as conventionally applied in psychological studies, may provide clearer insights into the mechanistic basis of decision making and information processing in behavioural ecology. This article is part of the theme issue 'Signal detection theory in recognition systems: from evolving models to experimental tests'.

The signal detection problem of aposematic prey revisited: integrating prior social and personal experience

Ever since Alfred R. Wallace suggested brightly coloured, toxic insects warn predators about their unprofitability, evolutionary biologists have searched for an explanation of how these aposematic prey evolve and are maintained in natural populations. Understanding how predators learn about this widespread prey defence is fundamental to addressing the problem, yet individuals differ in their foraging decisions and the predominant application of associative learning theory largely ignores predators' foraging context. Here we revisit the suggestion made 15 years ago that signal detection theory provides a useful framework to model predator learning by emphasizing the integration of prior information into predation decisions. Using multiple experiments where we modified the availability of social information using video playback, we show that personal information (sampling aposematic prey) improves how predators (great tits, Parus major) discriminate between novel aposematic and cryptic prey. However, this relationship was not linear and beyond a certain point personal encounters with aposematic prey were no longer informative for prey discrimination. Social information about prey unpalatability reduced attacks on aposematic prey across learning trials, but it did not influence the relationship between personal sampling and discrimination. Our results suggest therefore that acquiring social information does not influence the value of personal information, but more experiments are needed to manipulate pay-offs and disentangle whether information sources affect response thresholds or change discrimination. This article is part of the theme issue 'Signal detection theory in recognition systems: from evolving models to experimental tests'.

The Cognitive Ecology of Stimulus Ambiguity: A Predator-Prey Perspective

Organisms face the cognitive challenge of making decisions based on imperfect information. Predators and prey, in particular, are confronted with ambiguous stimuli when foraging and avoiding attacks. These challenges are accentuated by variation imposed by environmental, physiological, and cognitive factors. While the cognitive factors influencing perceived ambiguity are often assumed to be fixed, contemporary findings reveal that perceived ambiguity is instead the dynamic outcome of interactive cognitive processes. Here, we present a framework that integrates recent advances in neurophysiology and sensory ecology with a classic decision-making model, signal detection theory (SDT), to understand the cognitive mechanisms that shape perceived stimulus ambiguity in predators and prey. Since stimulus ambiguity is pervasive, the framework discussed here provides insights that extend into nonforaging contexts.

How cognitive biases select for imperfect mimicry: a study of asymmetry in learning with bumblebees

Imperfect mimicry presents a paradox of incomplete adaptation - intuitively, closer resemblance should improve performance. Receiver psychology can often explain why mimetic signals do not always evolve to match those of their models. Here, we explored the influence of a pervasive and powerful cognitive bias where associative learning depends upon an asymmetric interaction between the cue (stimulus) and consequence (reinforcer), such as in rats, which will associate light and tone with shock, and taste with nausea, but not the converse. Can such biases alter selection for mimicry? We designed an artificial mimicry system where bees foraged on artificial flowers, so that colours could be switched between rewarding or aversive. We found that when the colour blue was paired with a sucrose reward, other cues were ignored, but not when blue was paired with aversive compounds. We also tested the hypothesis that costs of errors affect how receivers sample imperfect mimics. However, costs of errors did not affect bee visits to imperfect mimics in our study. We propose a novel hypothesis for imperfect mimicry, in which the pairing between specific cues and reinforcers allows an imperfect mimic to resemble multiple models simultaneously. Generally, our results emphasize the importance of receiver psychology for the evolution of signal complexity and specificity.

The perfection of mimicry: an information approach

We consider why imperfect deceptive mimics can persist when it appears to be in the predator's interest to discriminate finely between mimics and their models. One theory is that a receiver will accept being duped if the model and mimic overlap in appearance and the relative costs of attacking the model are high. However, a more fundamental explanation for the difficulty of discrimination is not based on perceptual uncertainty, but simply based on a lack of information. In particular, predators in the process of learning may cease sampling imperfect mimics entirely because the immediate pay-off and future value of information is low, allowing such mimics to persist. This outcome will be particularly likely when the model is relatively costly to attack and/or the discriminative rules the predator has to learn are complex. Information limitations neatly explain why predators tend to adopt discriminative rules based on single traits (such as stripe colour), rather than on combinations of traits (such as stripe order). They also explain why predators utilize certain salient discriminative traits while ignoring equally informative ones (a phenomenon known as overshadowing), and why imperfect mimics may be more common in phenotypically diverse prey communities.This article is part of the themed issue 'Animal coloration: production, perception, function and application'.