shadow-minimization

There are at least four ways to minimize the shadow cast upon the substrate, and the first three minimize shadow on the ventrum as well. (1) The animal may avoid loci of strongly directional light, for example, by remaining in the shadow of other objects or by frequenting shadow-less environments such as sufficiently deep water where scattering eliminates directionality (ch 3). Thus, deep woodland birds and many fish cast virtually no detectable shadows because of directed-light avoidance.

2) Laterally compressed animals cast very small shadows on the substrate and almost none on themselves, and hence also have little problem being detected by shadow-contrast. Many fishes are so shaped, as are many insects; other insects at rest fold their wings vertically above their bodies (e.g., butterflies and damselflies) to create lateral body-compression.

3) By remaining closely appressed to the substrate an animal eliminates substrate-shadow, and if flattened dorso-ventrally eliminates or greatly minimizes shadow on its ventrum. A typical example is the moth that appresses itself closely to the bark of a tree. Flounders (Bothidae and Soleidae) are an extreme case, in which the compression may be due to other factors as well. Of course, lateral compression has the effect of dorso-ventral compression if the animal lies on its side, as do many flatfishes. My own observations on shorebirds on the open beach and mudflats suggest that when resting they often lie on the substrate instead of standing like many other birds, possibly an example of this tactic of substrate-appression.

Finally, (4) an animal can eliminate substrate shadow by simply remaining far from a substrate (fig 3-12, p. 77). However, to eliminate its umbra totally an animal must be more than 100 body-lengths from the substrate (eq 3.11), although its shadow will decrease in size as a linear function of the animal^’s distance from the substate below (eq 3.12). Insects can fly high enough to avoid casting shadows on the earth's surface, but ordinary (non-migratory) flight of most birds is below 100 body-lengths in altitude. It seems likely that because of scattering (ch 3) the umbral shadows cast under water are less distinct than in air, so that fishes and other aquatic animals may be able to remain well within 100 body-lengths of the bottom without casting definite body shadows. Furthermore, surface wave-action is liable to disrupt shadows in shallow water, so that shadows in aquatic environments will probably rarely be important. In general, then, substrate-avoidance may be a limited behavioral mechanism for avoiding tell-tale shadows.

In sum, substrate-shadows may be a relatively minor problem among animals. The shadows are most important when the animal is large in body-size relative to the irregularities of the substrate upon which its shadow falls, and here only in direct sunlight. In this cases, substrate-appression may be expected to be most common, and when the animal's dorsal coloration matches that of the substrate such behavior also makes its profile less detectable (see below).

counters-shading

A.H. Thayer seems to have originally proposed about 1896 that contrast between dorsum and ventrum due to the animal's own shadow could be diminished by counter-Shading, in which the dorsum has a lower reflectivity than the ventrum. The idea was developed by his grandson, G.H. Thayer (1918), discussed extensively by Cott (1957) and shown conclusively by De Ruiter (1958) and others to promote concealment of caterpillars from their predators.

A few considerations suggest where counter-shading should be effective: (i) the animal's body must be shaped so as to cast a shadow on itself, and (ii) its illumination must come habitually from above. These conditions are so readily met by terrestrial animals that "normal" coloration incorporates counter-shading in many insects, amphibians, reptiles, birds and mammals. Indeed, it is the exceptions that arouse curiosity, such as the conspicuous reversed counter-shading of the breeding male bobolink, which is tan and white above but jet black below (see ch 7) . Caterpillars that frequent the undersides of leaves and twigs are also reverse-countershaded, but their upside-down posture renders them truly counter-shaded in the environment. A fascinating example from fishes is the upside-down catfish, which true to its name swims upside-down, putting its dark belly upward (Sterba, 1962: 409 and plate 95).

For cylindrically shaped bodies, one expects the counter-shading pattern to be darker above the lateral midline, but in other cases the pattern may be different. If the animal is roughly rectangular in cross-section, only its belly is in shadow, and counter-shading is similarly patterned. For example, the white-tailed deer is brown to gray on its back and deep vertical sides, being white only on the belly. The exact pattern of counter-shading must therefore be examined in each case with regard to the animal's shape and the shadow it casts on its own parts.

The strength of the contrast between dorsum and ventrum in counter-shading should be a function of the degree to which illumination is directed from above. For animals living in open sunlight, one expects and finds strong contrast, but in animals adapted to deep woodland conditions or in open waters having much scattered light, the contrast may be more subdued. We expect counter-shaded fish only in relatively shallow water, and the coney is an excellent example. This species has been collected in many color phases, some of which are uniformly yellow to brownish red, but one of which is strikingly counter-shaded. Chaplin (1972: 14) notes that this latter is a "common shallow water form" and Bhӧlke and Chaplin (1968: 264) note that "this phase has been termed the 'excitement phase' and thought to occur when the fish is alarmed or is stimulated by the presence of food, but we have observed it as a longterm phase shown by fish cruising over the reefs." Figure 6-1 compares counter-shading with other mechanisms for suppressing contrasts due to shadows.

A few notes might be added to the subject of counter-shading. First, other factors--such as protection from UV-radiation (ch 4)--predict the same pattern of coloration, so that the counter-shading explanation should always be treated only as one initial hypothesis in any specific case. Second, reversed counter-shading leads to visual conspicuousness, and therefore may be used in social communication as a signal (ch 7). Finally, it is interesting to note a special kind of "counter-shading." Some bioluminescent fishes utilize ventral photophores to make the lower part of the body about as bright as the upper part (Hastings, 1971). This is only one of many uses to which bioluminescence is put by animals (see Lloyd, 1977). The types of suppression of shadow-contrasts used by animals are summarized in Table 6-I the following page.

Fig 6-1. Suppression of shadow-contrasts exemplified by various fishes. Substrate-avoidance suppresses only substrate shadows and counter-shading suppresses only body-shadows, but the other three mechanisms suppress both kinds of shadows. Some animals may exhibit more than one mechanism simultaneously.

Table 6-I

Types of Suppression of Shadow-contrasts

disruptive coloration

The outline of an animal may be disrupted by presenting a different outline that is visually more compelling. If the alternative outline resembles some recognizable shape, the deception works by confusing the observer's recognition (see initation, below). This section concerns deception that simply disrupts the tell-tale outline and hence diminishes the probability of the observer's detecting the animal against its background.

The visual principles of such disruptive (or ruptive) coloration appear to be unspecified, so it is worth attempting to identify some of them. (1) It seems likely that the contrast between surface patches of the object must be at least as great as the contrast between a patch and the background. Restricting the case to a two-color, achromatic animal, the ideal relation seems to be

where R₁ and R₂ are the reflectances of the patches and R_b is the reflectance of the background. Any other relation will render the contrast between at least one patch and the background stronger than the contrast between the two patches. Therefore, ideal disruptive coloration might be black and white, or at least dark brown and light yellow.

2) Unless the pattern deceives by resembling something else (see Imitation , below) , the patches should be randomly shaped, or at least irregular with respect to the animal's true shape.

3) The patches must be arranged such that light and dark alternate on the periphery of the animal as seen by the observer. Otherwise a light edge or dark edge outlines the animal and makes it more conspicuous (Cott, 1957).

4) It seems likely that if the patches are too small the animal looks patterned and may even look uniformly colored at a distance because of spatial mosaic fusion (p. 144). Similarly, if the patches are too large, the animal's real shape may be evident simply because the amount of deceptive contrast is so limited. Therefore, there should be an ideal size of disruptive patches, relative both to the distance at which the object is viewed and its own size. T. Johnston (pers. comm.) suggests further that the ideal disruptive pattern should match the "graininess" of the background against which it is seen.

Many patterns cited as disruptive also involve other principles of visual deception. For example, military camouflage of tanks and gun-implacements often involve matching vegetation color (see matching coloration, below) as well as disrupting tell-tale shapes. During World War II, warships were painted in disruptive patterns. These have since largely been abandoned, perhaps because the patterns were not effective deception, or more likely because with the advent of radar visual detection of ships is no longer a consideration.

Ostensible disruptive coloration of animal may often (like military camouflage) involve other principles, but convincing examples of disruption are common (Cott, 1957). Most, like the breast-band on many small plovers (Charadrius), are black and white, conforming to the first principle deduced above. The orange dog caterpillar of the giant swallowtail butterfly is dark brown and pale yellow, the patches being quite irregular in shape, thus conforming to the second principle. Huxley (1958) suggested that the killdeer, which is an exceptionally large plover, possesses two black breast bands because of its size (third principle). The breast-bands of plovers are not, however, irregularly shaped, nor are the markings of many boldly patterned fishes, so one suspects in such cases additional principles of deception are being combined with disruptive coloration. G.W. Barlow (1967) has suggested that the middorsal stripe on the head of certain fishes is disruptive coloration: the "split-head" pattern.

transparency

Perhaps the most straightforward solution to suppressing an animal's outline is to eliminate it by making the animal transparent so the observer sees the background through the animal. Certain optical and morphological considerations suggest where animal transparency is most likely to be found. It is not sufficient for an animal simply to transmit light: even highly transparent substances such as glass refract (ch 3) because they are optically denser than the medium. The distortion due to refraction of an animal may be minimized in at least two different ways: either the animal evolves an index of refraction close to that of its medium, or else it remains very near a background so that visual distortion is minimized. Both strategies appear to be used, at least sparingly, for visual deception.

All animals, because they are constructed of macromolecules with high molecular weight, probably have relatively large indices of refraction. Highly transmitting organic substance such as oils and fats have refractive indices generally between 1.4 and 1.5. Therefore, it seems almost impossible for an animal living in air (having a refractive index near unity) not to refract light. However, water (n = 1.32 to 1.33, depending upon temperature) provides a suitable medium, providing the animal has a high enough water content to bring its own refractive index reasonably close to that of the medium. It is not surprising, then, that most transparent animals are aquatic, familiar examples being certain coelenterates, ctenophores and many smaller animals that drift in the ocean, as well as some adult fishes and fry, and a few other freshwater animals. Of animals living in air, the wings of dragonflies approach transparency, but similar examples are relatively few.

There is a second optical principle of transparency that may be utilized by certain aquatic animals. As pointed out by Ruechardt (1958: 30-31), an object that has a different index of refraction from its medium can still be invisible if illuminated homogeneously. He outlines a simple experiment whereby a transparent glass rod may be rendered invisible in air--an experiment that astounds even those familiar with optical principles. Place the rod inside an inverted cone painted highly reflecting white inside, and illuminate from above; by peeking through a pinhole in the cone one looks at the rod nearly homogeneously illuminated and finds it difficult to detect. Spatially homogeneous illumination does not occur in terrestrial environments, but in deep water illumination approaches this ideal (ch 3) so that transparent animals with indices of refraction different from that of water may still become nearly invisible.

Transparent animals living in air must adopt the strategy of remaining very near some background, since the highly refractive animal would be easily detected in free space. J. Baylis and E.H. Burtt (pers. comm.) have pointed out that the wings of certain dragonflies (Suborder Anisoptera) and clear-winged moths (Aegeriidae)--which are held horizontally while at rest--are sufficiently transparent to be visually undetected. Wings of many species have dark markings that are thus visually disconnected from the resting animal, but become visually conspicuous when they flicker in flight. It should be pointed out, however, that wings of the related damselflies (Suborder Zygoptera) are held dorsally over the resting animal despite being similarly transparent (but see shadow-minimization above).

It is also useful to note that animals may be almost entirely transparent, as ctenophores, or have only certain transparent parts, as in the fins of certain fishes. By being partially transparent an animal may alter its shape as perceived by another, and therefore may go unrecognized For example, a leaf fish with transparent fins does not have the visible shape of a fish, and hence may go unrecognized by predators and prey alike. Finally, transparency is not always due to selection for visual deception. For example, certain bathypelagic fishes are transparent, but live in environments where light levels are effectively zero. It seems reasonable to assume that the metabolic cost of producing unnecessary pigment is sufficiently high to be selected against in animals living in poor trophic environments.

matching coloration

The third solution to suppressing the visual outline of an animal is for its perceived coloration to match that of its background. There are actually two ways of making the spectral radiance of an animal match that of its background: by achromatic (neutral) reflectance and by spectral reflectance similar to that of the background's.

My notion of "spectral conformity" of achromatic reflection comes from noticing the widespread occurrence of white ventral ground coloration in birds. Although light ventral coloration is a factor in counter-shading (see above), it need not be white to be effective. White coloration, being a selectively neutral reflector, will tend to take on the hue of ambient irradiance and under specified conditions will also match the background hue. Two conditions must be met: (a) the reflectivity of the animal must be neutral (achromatic) and relatively high, or else the animal will be colored (chromatic reflectivity) or gray to black (low reflectivity); and (b) the ambient irradiance reflected from the animal must resemble in spectral distribution the ambient radiance emanating from the background. Put differently, the second condition means that the light must be colored by its having been reflected or scattered from the background, or else filtered in such a way that the filtered light has the same spectral distribution as that emanating from the background.

One common example fulfilling both conditions is the white underparts of many forest songbirds. The ambient illumination is greenish because of transmission by canopy leaves, and light reflected from lower leaves is even greener in spectral distribution. (Not all substances transmit and reflect the same spectrum, but leaves do.) Therefore, the light striking the white breast will be greenish in spectral composition, and hence the breast will reflect the greenish emphasis. Because the reflecting leaves are receiving approximately this same irradiance, and then further absorbing, they will appear both darker than the white bird and of a more saturated hue. Therefore, the deceptive effect is subtle and probably useful only in conjunction with other principles of visual deception, such as counter-shading.

The same principle should apply to animals that frequent a homogeneous substrate in open sunlight, if they remain close to the substrate so that their undersides are irradiated primarily by reflected light. The white breasts of small sandpipers and plovers thus utilize such reflectance from sand and mud, whereas longer-legged relatives may stand too high off the ground to be illuminated primarily by substrate reflectance. (For example, the curlews, Numenius spp., tend to be quite dark below.)

It seems likely that white reflectivity is also a factor in the neutral reflectances of so many silvery colored fishes. Open water species, unless they are near the surface, will be irradiated largely by bluish light scattered by the water. By reflecting this light achromatically, the fish blends well into the background of scattered light.

An obvious way to blend into the background is to match it in spectral reflectance characteristics. Such chromatic reflectance may involve one homogeneous reflection spectrum (uniform chromatic reflectance) or a patterned reflective surface resembling the background (patterned chromatic matching).

Uniform chromatic reflectance occurs when a homogeneously reflecting surface has the same spectral reflectivity as its common background, so the advantage of such a deceptive strategy over achromatic reflectivity is that the match can be quite convincing visually, but the disadvantage is that the match works only against one background. Uniform chromatic reflectance is so widespread and frequently documented (e.g., Cott, 1957; Wickler, 1968) that only a few examples need be pointed out. Many arctic and alpine animals that live habitually on or near snow are permanently white (e.g., the mountain goat and snowy owl) whereas others are white in winter but turn a matching brown in summer when the snow is gone (e.g., the snowshoe hare and white-tailed ptarmigan). Dice (1947) was the first to show experimentally that predators find prey more readily when the latter contrast in brightness with their substrate. Perhaps the most elegant and well-documented uniform matching is the case of industrial melanism, in which the European peppered moth living in areas where carbon deposits cover the environment have become melanistic (e.g., Ford, 1964; Kettlewell, 1973). Uniform chromatic reflectance is commonly utilized in conjunction with other deceptive principles, as well as dictating the entire body coloration. Industrial melanism also occurs in North American moths (Owen, 1961).

One study of matching uniform chromatic reflectance impresses me as important in showing the way toward future studies. Johnson and Brush (1972) have studied the polymorphic coloration of the Central American sooty-capped bush tanager. First, the coloration itself is documented objectively with reflection spectrophotometery (ch 3). Second, the C.I.E. chromaticity variables (table 5-I) are used to describe the visual appearance of the birds, and then used analytically to establish the pigmentational basis. The darkness of the tanagers is due to concentration of the biochrome (lutein), and as the concentration increases the excitation purity increases linearly with it, indicating a single-pigment basis for coloration. Geographic variation is then used to show that the dark gray-green morphs are restricted to high volcanic mountains where "the background of ash, vegetation, and fog provides an environment of pervasive grayness." Employing the comparative method (table l-III) , Johnson and Brush find many other avian species with gray plumage in this habitat, and when the species range widely into other areas they are most common in the gray environment. Finally, as a tantalizing footnote, there is some evidence that the gray-green phase of the tanager increased in abundance following the major eruption of Volcan Irazu in 1963. This kind of careful study sets standards for future research on animal coloration.

Some caution is necessary in identifying matches in chromatic reflectance. We humans tend to classify everything according to our own sensory systems, but Eisner et al . (1969) provide a clear example of problems with this. Crab spiders, which frequent floral heads, may match the flower coloration to our eyes and hence be concealing to predators such as birds. However, when viewed with an ultraviolet analyzer, they may be quite conspicuous and hence could be easily detected by their insect prey having good UV-sensitivity (ch 5).

A special kind of uniform chromatic reflectance seems worthy of mention because it appears to be widespread. Many insectivorous birds have dark lines or patches around the eyes, which themselves have dark irises, so that the tell-tale outline of the eye is not revealed. The eye is in a black field in many woodpeckers, flycatchers, swallows, corvids, chickadees and so on, although other explanations for black about eyes are possible (ch 4) and no critical studies distinguishing the hypotheses seem to exist (also see discussion in Burtt, 1977).

When the substrate is patterned rather than being uniformly reflecting the animal may evolve patterned chromatic reflectivity to match it. Many ostensible cases may involve disruptive principles rather than strict matching, but there are convincing examples. Wickler (1968: 55, figure 6) shows a cogent photograph of the sole against a sandgrain background, and many flatfishes (flounders, Bothidae; soles, Soleidae; and tonguefishes, Cynoglossidae) resemble or can change color to resemble their patterned substrates. The light form of the peppered moth in non-industrial areas is virtually invisible on the patterned bark of oaks (see photographs on p. 91 in Bishop and Cook, 1975). G.W. Barlow (1974: 27) points out that the surgeon fish Acanthurus guttatus, which is found in the surf with many surface waterbubble, is marked with numerous white spots. Sometimes variation within a species yields a correlation between habitat-choice and concealing coloration. Schoener and Schoener (1976) found that small female brown lizards perch on small branches and have longitudinal stripes whereas larger females perch elsewhere and are unstriped.

In some cases, special matching structures may accompany matching patterns. For example, some small inshore bottom fishes of California show blotchy patterns complete with coloration that looks like encrusting algae. The one spot fringehead is named for the cirri projecting dorsally from its head, the cirri resembling algal filaments common in its environment (see Miller et al., 1965: 57). Other clinids and cottids have similar coloration and structures. Such matching structures grade into element-imitation, considered below.

At least some flounders (Bothus) help further to suppress their visual outlines by flipping sand along their margins while lying appressed to the substrate. This matching strategy may be called material matching to indicate that the animal uses materials from its environment. Classical examples are crabs of various species that place bits of their environments on top of their carapaces; algae, anemones and various other plants and animals often thrive on the crabs. Wickler (1968: 56) provides other examples. Material matching for concealment may be difficult to distinguish from simple physical shelter, and hermit crabs (which live in gastropod shells) are a case in point.

Fig 6-2. Suppression of outline-contrast by various mechanisms.

Examples of these strategies for suppression of outline contrast--including disruptive coloration, transparency and various forms of matching coloration--are shown in fig 6-2 (opposite). The principles of suppression of outline are summarized in table 6-II above.

Table 6-II

Suppression of outline-contrast

concealing imitation

Whenever an animal is evolved to resemble something it is not, it could be said to mimic the other object. However, "mimicry" is used both in this broad sense and also in the restricted sense of imitating another species of animal. To avoid confusion, I restrict the term to the latter sense (calling the phenomenon animal-mimicry for further clarity) and refer to other imitation as concealing imitation . In concealing imitation, an animal resembles some particular kind of object normally found in its environment, and hence is distinguished from matching coloration (above) in which the animal merely suppresses its tell-tale outline. Both deceptions conceal the animal because it is overlooked as an irrelevant part of the environment.

Two kinds of concealing imitation may be distinguished: element-imitation and object-imitation. In element-imitation an animal resembles some specific and common object that is one element of an environmental pattern. In patterned chromatic reflectance (above) the animal duplicates over its surface the pattern of the environment, but in element-imitation the entire animal becomes just one element of the visually patterned environment. The principal examples of element-imitation are found among insects, where shape as well as the coloration of the animal imitates some element of its environment. Walkingsticks (Family Phasmatidae) and other orthopterans resemble the twigs of trees and bushes they feed upon, the pupae or larvae of gossamer-winged butterflies (F. Lycaenidae) resemble tree buds, treehoppers (F. Membracidae) and other homopterans resemble thorns on their host plants, and so on. Wickler (1968) provides many examples of element-imitations (esp. pp. 50 and 60-66).

Ordinarily, behavior, physical structure and surface coloration go hand-in-hand to produce an effective imitation. Insects are not only colored and shaped to resemble parts of plants, but the animals also seek loci that are appropriate to the deception. M. Itzkowitz pointed out to me in Jamaica the habit of the elongate trumpetfish, which swims head-downward among staghorn and other elongate corals thereby becoming remarkably concealed. In this case, it is uncertain whether the shape and coloration of the animal are selected to be imitative, but the behavior certainly appears to be.

In object-imitation , the animal resembles some specific object that is not necessarily common and is not an element of a regular environmental pattern. Object-imitation works not because the imitator blends into the background pattern, but because the viewer perceptually classifies it as something it is not. There is probably a continuum between element- and object-imitation, but the major distinction drawn here is that element-imitators succeed primarily when they are among the elements they resemble, whereas object-imitators succeed in deception even when no examples of the object they imitate occur in the same visual field of the viewer. A typical example is the predatory leaf fishes (Nanidae) floating by in the water; dead leaves are also imitated by many insects ( e.g., among butterflies, adult Anoea and Polygonia , pupal Limenitis and Polygonia and larval Limenitis ; Klots, 1951). Cephalopods may hold out their arms to resemble drifting Sargassum (Moynihan, 1975). An outlandish example of object-imitation is the young Papilio and Limenitis larvae that resemble bird-droppings. An avian example, which may bridge the gap between element- and object-imitation, is the great potoo of Latin America, which sits motionless in an unbird-like position on dead stumps, looking much like the terminal fragment.

There is still another kind of imitation that represents some sort of transition between resembling one element of a pattern and a unique object. Wickler (1968: 60-64) draws attention to imitating groups of animals. Each cicada of an African species in the genus Ityraea resembles a flower, but the individual need not perch among flowers to provide deception, since the insects are social and whole groups resemble inflorescence even when no real flowers are present. Thus, each individual is an element-imitator and the entire group is an object-imitator. Wickler (loc. cit.) also recounts the extraordinary behavior of congregating Tubifex- like worms that together look like a sea anemone.

For imitation to be effective, the deceived animal should have no great interest in the thing imitated. Furthermore, it is necessary that the deceived animal not be able to make a connection between the imitator and imitated that would be useful in searching for the latter. If all stumps have potoos sitting on them, stumps become a clue to finding potoos, so to be effective stumps without potoos should outnumber those with potoos.

animal-mimicry

When an animal imitates another species it is usually visually conspicuous (easily detected in the environment) but tends to be classified incorrectly by an observer. There is probably no firm logical distinction between object-imitation and animal-mimicry, but most cases fall naturally around one or the other of these points on a continuum. Animal-mimicry is therefore a special kind of object-imitation in which the "model" after which the mimic is patterned is another species of animal. A common kind of animal-mimicry is the "fishing lures" used by some predators, which have modified parts that resemble the prey of a species upon which they themselves feed. The alligator snapping turtle has an elongated dorsal projection from the floor of its mouth that is pink and wriggles like a worm; fish coming to investigate are quickly snapped up by this bottom-burrowing reptile. Analogous lures are dangled above the mouths of sit-and-wait predatory fish such as frogfishes (Antennariidae) and batfishes (Ogcocephalidae).

There are many examples besides lures for predation in which one part of an animal mimics another animal. Wickler (1968: 137, figure 30) illustrates an elaboration of the mantle of the clam Lampsilis ovata ventricosa that looks like a small fish. When predatory fish snap at this lure, the female clam squirts its larvae into the fish's mouth, where they develop as parasites. Hence mimetic lures are not used only for predation. An oft-cited example of mimicry is the eye-like spots adorning all kinds of animals (see Wickler, 1968 for many examples). When these spots occur in pairs, as on the underwings of some moths, and when it is shown by behavioral experiments that they frighten potential predators such as birds (Blest, 1957), they are convincing cases of eye-mimics . Similar paired spots are displayed by cephalopods (Moynihan, 1975). However, many animals are covered with spots of various kinds, and some investigators have glibly assumed these are eye-mimics without any real evidence. This problem is taken up again in ch 7.

Entire animals may also mimic other kinds of animals, and examples are common. One species of spider I have photographed (probably Synemosyna lunata) is remarkably similar to an ant, and walks around with its forelegs raised like antennae, giving the impression that it is six-legged. Orchids are famous for their flowers that resemble the hymenopterans that pollinate them, so that animal-mimicry extends even across the kingdoms. Wickler (1968) relates the grisly strategy of a goby that so strikingly resembles a cleaner-wrasse it may approach a large fish with impunity and then take a bite from its flesh. Wicklerꞌs book should be perused for many other extraordinary examples of animal-mimicry.

Most examples of animal-mimicry involve similar animals, such as one insect resembling another or one fish looking like a different fish species. However, Huey and Pianka (1977) report a juvenile lizard that appears to mimic a noxious carabid beetle not only in size and coloration, but also by body posture and locomotory movement. Yet one must be ever aware of convergent evolution in postulating mimicry. It may be, as Eaton (1976) suggests, that the cheetah kit mimics the aggressive honey badger. However, the honey badger uses reversed counter-shading to be conspicuous (see next chapter) to potential predators, whereas the cheetah kit could be using the same pattern to be conspicuous to its mother or for some other reason. If the kitten lies motionless while the mother is away, its dark ventrum does not show and the young animal is concealingly colored because of matching its background. The biological literature is replete with examples of ostensible animal-mimicry, but in many cases not enough is known to be relatively sure that mimicry is really the basis of the convergence in color pattern.

Batesian mimicry is a subset of animal-mimicry, based on predator-prey interactions. This term has been applied to those mimics whose models are distasteful, poisonous, toxic or otherwise inedible or dangerous to the predator. The classical example is the edible viceroy butterfly that closely mimics the more plentiful monarch, which birds find distasteful. No discussion of mimicry is complete without mention of warning (or aposematic) coloration displayed by animals that could be harmful to a potential predator. They escape predation not by visual deception, but by "honestly" communicating the fact that they can sting or otherwise be undesirable to other animals. Efficiency, and hence effectiveness, of such interspecific communication is enhanced when different species display the same kind of warning coloration so that they are easily recognized by predators. This evolutionary convergence to a common type, in which every species is both model and mimic, is called Millerian mimicry . The black-and-yellow coloration of stinging hymenopterans is a typical example, but they in turn are mimicked by harmless dipterans, so that both Mllllerian and Batesian mimicry may occur in one evolutionary complex. A somewhat analogous example occurs among cleaner fish that have evolved similar yellow and black color patterns for recognition by their hosts, and in turn the pattern is deceptively mimicked by the non-cleaner goby mentioned above. Because these fishes are not involved in classical predator-prey relations, the example would not be called Batesian and Mullerian mimicry. Also, see Wickler (1968) for the difficulties in separating these two forms of mimicry.

A special case of animal-mimicry is behavioral mimicry , in which one species acts like another without necessarily being morphologically similar. The primary example of purely behavioral mimicry appears to be the "rodent-run" actions of some ground-nesting birds, which dart out and run along the ground like a mouse to distract predators away from the nest-site. Ornithologists classify this behavior functionally as a "distraction display" (Armstrong, 1964) which also covers deception that is differently based (see next section). To return to the cleaner-mimic story, the goby mentioned above also swims like the wrasse it resembles in coloration. The eastern indigo snake flattens its head and neck vertically, hisses and vibrates its tail to produce a rattling sound when cornered. This behavioral mimicry of a rattlesnake presumably is a defensive mechanism, but it is interesting to note that the indigo snake is known to eat rattlers (Conant, 1975: 187).

Finally, it is worth pointing out animal-mimicry that depends entirely upon emission of an optical signal. Female Photuris fireflies mimic the mating signals of female fireflies of other species, attracting the males, which they seize and devour (see Lloyd, 1977).

feigning

imitative behavior may be part of both concealing imitation and animal-mimicry, but when the animal imitates its own behavior one may refer to it as feigning. In feigning, the observer both detects and recognizes the animal, but may mistake its behavior.

Fig 6-3. Examples of imitation, including the spectrum of concealing imitation (above) with the treehopper resembling a repetitive element in its environment (thorn) and the leaf-fish resembling an isolated object (leaf). The imitations are often more compelling in nature than shown here to emphasize the animal itself; for example, the transparent fins of the leaf-fish are virtually invisible, rendering its body shape less fish-like and more leaf-like. The ant-mimic (below) is the spider Micaria and the opossum is feigning death.

Table 6-III
Types of Imitation

Feigning is a common form of deception used to distract predators from the nest (Armstrong, 1964) and hence can function like behavioral animal-mimicry. The classicail case of nest-distraction is the injury-feigning or "broken-wing act" of the killdeer, in which the incubating bird runs quietly to a meter or so from the nest when a predator approaches, and then lowers one wing and drags its body in an irregular motion in a direction away from the nest. The deception distracts the predator, which may even try to catch the parent, only to have it fly off.

Injury-feigning is not the only type of behavioral deception; death-feigning is also common (e.g., Franq, 1969). The Virginia opossum is so famous for going limp when approached by a predator that its behavior has given rise to the American colloquialism playin' 'possum . Indeed, since movement is such a widespread characteristic of animals, motionlessness per se is a deceptive mechanism promoting concealment, and might be classified as a special kind of feigning. "Freezing" to remain concealed is so common a ploy, especially among young and small animals, that specific examples are unnecessary.

Types of imitation are illustrated in fig 6-3 on p.179 and the optical principles upon which they depend are summarized in table 6-III (opposite).

symmetry-deception

Often an observer may predict something important from the symmetry of the animal it sees. For example, one expects an insect to jump away anteriorly. By creating an ambiguity concerning the symmetry, the animal deceives the observer. It is evident that symmetry-deception is not far removed from animal-mimicry, except in this case one part of the animal mimics another, rather than one species mimicking another. It is also the morphological analog of feigning, in which the animal's behavior mimics another of its behavioral patterns, once more emphasizing the spectrum of deceptive strategies and their relationships.

Wickler (1968) provides several examples of symmetrydeception among insects, a particularly compelling one being an unidentified lantern-fly from Thailand (p. 73, figure 14). Such ambiguity is not restricted to insects, however, since some snakes have colored tails that appear as heads, and this kind of deception should be looked for among other animals. The way in which butterflies are displayed in museum collections and pictured in books--with wings spread and flattened--draws our attention away from symmetry-deception when the wings are folded above the animals in their usual resting posture. Wickler (p. 76, figure 15) pictures Theola togarna at rest, with its wing projections looking very much like head and antennae at the rear of the animal. This butterfly also turns around upon landing, so that the symmetry-deception is enhanced by the behavior. It might prove interesting to look at the landing behavior of swallowtails, hairstreaks and other butterflies with posterior projections on their hindwings.

startle-deception

Some symmetry-deception seems to benefit the animal primarily by rendering its anticipated direction of movement ambiguous, and another way in which this same end is achieved is by visually startling the observer just before a change in direct. Such startle-deception seems to function by creating a visual change at the instant of directional change. However, startle-deception seems not to be well studied, and might be a heterogeneous category of some complexity, as two examples illustrate.

Many butterflies--Polygonia, Cereyonis, Vanessa, Eunica, Anea--have bright and colorful dorsal surfaces on the wings but dull undersides. Their flashy colors are exposed in flight, but then they drop into vegetation suddenly and become nearly undetectable (Klots, 1951) by holding their wings dorsally (see shadow-minimization and matching coloration , above). In this case, it appears as if a pursuing predator or other viewer might develop a perceptual set (see next section) based on the exposure to the bright coloration, and hence become confused when the pursued image no longer exists. Underwing moths and certain grasshoppers have similar behavior and coloration, and in grasshoppers a sudden sound contributes to the startle.

Ostensible startle-deception in silvery colored fishes may depend on a different effect. The mirror-like surface reflects light specularly (fig 3-5) , so that the fish is dull at most angles but reflects a dazzling light (ch 5) at the angle equal to that of the incident light. While diving in Florida, I have noticed that when a school of dull lookdowns turns suddenly, one sees many dazzling flashes of light and then it becomes difficult to find the fish immediately thereafter. This example of startledeception may work on combining several factors. The flashes constitute a visual surprise for which the viewer is unprepared, hence breaking the perceptual set and possibly frightening the viewer by means of an abrupt and not immediately recognizable stimulus. The timulus may be bright enough to dazzle the eye (ch 5). Then, the brightness of the stimulus may leave an after-image (ch 5) , or create a new perceptual set, either of which make it difficult to detect the now-dull fish an instant later. Finally, the fish--having changed direction at the time of the flash--are in an unanticipated part of the visual field, moving in an unanticipated direction because the turn itself was disguised visually by the flash. This example, based on my introspective experience in the role of potential predator, accords with observations by J. Baylis (pers. comm.) on other silvery fish, and also demonstrates how complex may be the basis of startle-deception.

An apparently analogous ploy is used by the bioluminescent flashlight fish (Morin et al., 1975. It swims οver the reef slowly with its light on, then blinks the light rapidly and changes direction with an increase in swimming speed. This is just one of many uses of bioluminescence in this fascinating fish.

replicate-deception

Wickler (1968: 57-59) has drawn attention to a ploy that allows long-term deception as to an animal's location: the animal constructs replicates of itself in the environment. Such replicate-deception is known in a spider that builds other "spiders" of its web-material and in a moth that severs several leaves to look like its own leaf-wrapped pupation-case. Replicate-deception works by saturating the environment with many similar things, and the ploy may be reversed by making similar things appear different, as discussed in the next section.

deceptive polymorphism

All species of animals exhibit variability among individuals to some degree, and when this variability is discontinuous the species is polymorphic. Polymorphic variation in coloration is sometimes called polychromatism, and many species are known to have two or more color morphs or phases that are not correlated with sex or age. The reasons for polymorphism are often obscure, as in the red and gray color phases of the screech owl. In some cases, however, an argument can be made for polymorphic coloration being an adaptation for visual deception of predators: deceptive polymorphism. A predator that successfully takes one prey animal may concentrate on searching for others that look like it, and hence overlook conspecific prey that are differently colored. Psychologists know this behavioral phenomenon under the rubric of perceptual set (one tends to see what one expects or wants to see). The ecologist L.Tinbergen (1946) wrote of the predator having a search images based on previous captures.

Polychromatism of the European snail Cepea is one case (B. Clarke, 1962). Presumably predators continue to hunt for one type that therefore becomes increasingly rare until rarity causes the predator to hunt randomly and discover a new type. In this way natural selection maintains the polymorphism. It seems possible that the extreme intraspecific variability of coloration found in other gastropods may be examples of deceptive polymorphism, such as in the common dove shell of shallow waters in Florida. Wickler (1968: 59-60) provides several other examples in caterpillars, moth larvae, snails and beetles.

Deceptive polymorphism is more closely related to startle-deception than one might at first think. Startle-deception depends upon rapidly changing appearance and polymorphism functions by different appearances of members of the same species. One can imagine these strategies combined in an animal that simply changes color from time to time in order to prevent any observer building up a perceptual set or seach image. Such a phenomenon might be looked for in animals capable of facile color-change, such as fishes and cephalopods.

Examples of visual ambiguity--including symmetry deception, startle-deception, replicate-deception and deceptive polymorphism--are shown in fig 6-4. There remains to be considered, though, one last type not illustrated.

Fig 6-4. Examples of visual ambiguity.

motion-deception

Jackson et al. (1976) analyzed color patterns of North America snakes and found a correlation between anti-predator strategy and pattern. Specifically, species that primarily flee, rather than defensively maintain their ground, tend to be marked with longitudinal stripes. The apparent optical advantage of this pattern is that it distracts attention from cues of movement, making it difficult for the predator to track the snake. Visually, the longitudinal stripes seem stationary while the snake moves forward. This principle of motion-deception also appears to be utilized by cephalopods (descriptions by Moynihan, 1975; interpretation mine). In behavior having "a strong escape component," the animals assume a posture with the arms straight out anteriorly and change into a color-pattern of bold longitudinal stripes. Another supporting case comes from open-water fishes which have no vegetation in which to hide. Due to biomechanical principles, such species tend to be long and thin (Alexander, 1967), but they also tend to be marked with longitudinal stripes, primarily for motion-deception I believe. There is a strong, albeit imperfect, correlation of habits, body-shape and pattern among fishes of coral reef areas, as thumbing a book such as Chaplin (1972) quickly confirms: those species that stay near the substrate tend to be deep-bodied with vertical bars and those that cruise over the reef in open water tend to be long with horizontal stripes. The principles of visual ambiguity are summarized in table 6-IV.

Table 6-IV

Tyes of Visual Ambigutity