biolumines cence

Animals generate their own light in one of three ways. Some have extracellular organs that secrete luminescent materials into the sea around them, as is known from a shrimp and from a squid that squirts a glowing cloud as part of its escape mechanism. Other animals secrete a luminescent Slime over their bodies, and this process involves more than simple dumping of chemicals into the water, although it has not been well studied. The third, and usual, mechanism is intracellular bioluminescence, in which light-producing cells are collected into a functional organ.

Photic organs are known from many of the major groups utilizing optical communication, including cephalopods, crustaceans, insects and fish. Such organs differ among species, but are usually much more than a collection of bioluminescent cells. The organs may have light-absorbing layers that can shield the luminescent structure, light-reflecting layers to enhance propagation of light away from the animal, filtering tissues to change the color of luminescence and refractive bodies resembling lenses, in addition to nerve supplies and in some cases movable covering structures. In fact, many such photic organs have the superficial appearance of an eye. The ultimate color of observed luminescence is determined by the molecular mechanism plus the reflecting layer and filters.

Not all light-emitting animals are actually bioluminescent themselves because some contain luminescent bacteria--the only case known to me where one symbiont of a pair derives purely communicational advantage from the relationship. It is often difficult to tell whether the animal or its symbionts are generating light without detailed anatomical investigations. Although most luminescent bacteria glow chronically, certain fishes can modulate the emitted light by moving a cover over their microorganisms.

The chemical mechanisms of bioluminescence are incompletely understood and undoubtedly vary somewhat among species. A pigment-like, heat-stable molecule of low molecular weight known generically as luciferin emits light when it is oxidized. The oxidation is mediated by a large, heat-labile, water-soluble catalyst, chemically identified in some cases as an enzymatic protein and hence known as luciferase. The reaction always requires water and oxygen, although the partial pressure of oxygen needed may be quite low, and in some cases other chemical constituents may be needed. Bioluminescence has a longer latent period than the radiation-induced processes of some fluorescence and phosphorescence (ch 3), and emits light over a longer time-span once activated.

The spectrum of bioluminescence usually has a bandwidth of between 80 and 200 nm, and always occurs in the visible spectrum from about 415 to 670 nm, often centered about 480-510 nm. A single animal may bioluminesce in two different colors; in some cases this is due to a single mechanism rather than two, depending upon elaborate anatomical specializations associated with the basic light-producing mechanism.

The temporal control of light-emission, which is the primary information-carrying variable in bioluminescence, has various sources. In some cases, the neurally controlled production or release of luciferin and luciferase determines when luminescence shall take place. In other cases, the necessary water or oxygen is controlled, and in some cases the metabolic machinery altering paths producing precursors or other necessary chemicals is the variable under physiological control. In addition, animals that possess an opaque covering for a bioluminescent structure are able to open and close that covering by conventional muscle contraction. All these factors lead one to expect good temporal control over light-emission, which seems to be the case in at least some well-studied species such as fireflies (e.g., Lloyd, 1977).

muscular movement

Almost all optical communication depends in some way upon muscular movement. Bioluminescent organisms may move in space while emitting light or modulate emitted light by means of movable covers (above). Other animals change color by chromatophoric mechanisms involving muscular contraction. Animals that create extrinsic optical signals such as the display bowers of bowerbirds do so by coordinated movements. Finally, animals use muscular movement to change their body shapes, orient in space and make gestures.

Animal movement is due to subcellular microtubules (ciliary and flagellar action) or myofibrils (muscle contraction). Anatomical details of muscular types, arrangements and innervation vary greatly among animals and likely impose various constraints on signaling that have yet to be studied. Vertebrate smooth muscle -responsible for mammalian pupillary responses, vasomotor changes as in blushing, and other optical signals--often has antagonistic sympathetic and parasympathetic innervation to the same muscle cell. Some cells are polyterminally innervated (many junctions), but many are not innervated at all, their activation being due to contraction of nearby fibers, mechanical stretching or free neuro-endocrine chemicals. Vertebrate smooth muscle is therefore likely to be the basis of relatively slowly modulated signals.

Skeletal muscles tend to be arranged in antagonistic pairs, as when one muscle mass extends a limb and the other flexes it. In crustaceans, which utilize limb movements as signals, muscle cells are innervated both polyterminally and polyneuronally (many different nerves), allowing for precise and rapid control. Insect muscles are similarly innervated, but lack peripheral inhibitory innervation; instead, inhibition occurs centrally to prevent activation of excitatory motoneurons. This difference is probably important in the kinds of signal movements used by the two arthropod groups, but the subject appears to be unstudied. Vertebrate striated skeletal muscle has monoterminal and mononeuronal innervation, the muscle cells themselves propagating an electrical spike potential that activates the entire fiber more or less synchronously and at a constant rate under given mechanical loading. It also appears as if there are slow- and fast- contracting fibers in the same gross muscle. Contraction of a single fiber is an all-or-none phenomenon, and about 10 to 1000 fibers innervated by the same nerve make a motor unit. Contraction is graded because a muscle may have hundreds of such units, only some of which are activated at a given time. Arthropod muscles are not organized into such units, their contraction velocity being due to the polyneuronal innervation of individual fibers that may contract slowly or rapidly. It seems possible that the "jerkiness" of signal movements in arthropods compared with vertebrate movements might be partially traceable to these differences in muscular mechanisms.

Because of these control mechanism, the temporal modulation of signal movements is not limited by contraction latency per se in vertebrate striated muscle. Hummingbird wings may beat at 50 Hz, for example. Some insect wings beat much faster--as high as 1000 Hz--a value that clearly exceeds the capabilities of muscles to contract and relax with each wing-stroke. The elastic exoskeleton to which the wings are attached is bent by muscle contraction that loads the system with potential energy this is released when it pops back into shape--much like the way in which a metal soft-drink can pops hack when the top is pushed down. The insect muscles then contract only occasionally to reinforce the oscillation--like pumping a child's swing on every third or fourth cycle. I have not found examples in which mechanical resonance systems are used in optical signaling, but clearly such rapid movement exceeds flickerperception rates under most conditions (ch 5) and hence is unlikely to be useful in communication. It seems likely to conclude that muscle mechanisms impose no serious temporal problems for animal signaling.

biochromes

Coloration of animal parts is created by molecules that absorb light differentially within the spectrum (biochromes) and specializations of the integument that affect light through physical principles such as refraction, scattering and so on (Schemochromes, below). The terms biochrome (D.L. Fox, 1953) or zoochrome (Needham, 1974) help prevent confusion due to the overworked word "pigment." In industry, a pigment is a coloring compound that is suspended in a base, such as pigments in paints, and a dye is a substance that complexes chemically with some substrate, such as dyes in cloth. Biologists, however, use the term pigment in a variety of ways, such as for the chromophore-protein complex in the eye that changes upon absorption of light (see visual pigments in ch 5) and for the chemical complexes of the integument that are photostable.

Biochromes in the integument are molecular complexes that absorb photons in particular spectral bands, turning the absorbed energy to heat (ch 3), and reflect non-absorbed photons to give animals their observed coloration (see eq 3.4). Molecules that absorb uniformly in the visible spectrum, so as to reflect white, gray or black, are also classified as biochromes. The molecular factors that determine the absorption bands of biochromes include their large molecular size, bonding structure of component atoms, planar configuration, side-chains and rings, and conjugation with proteins and other materials.

Because of the many ways in which the molecules can differ, there are hundreds, if not thousands of different animal biochromes. They tend to fall into about a dozen major classes based on chemical similarities (table 4-I), but most biochromes are incompletely characterized, so that the classification is likely to change. Many biochromes were first named before anything was known of their structure, then renamed to express general chemical similarities with other known biochromes, then renamed again when something of their chemical structure was understood. Even the major classes have various names, as indicated by comparisons in table 4-I. The approximate, if not complete, chemical structure is now known for at least on biochromes in each major group.

Table 4-I

Roughly Equivalent Classifications of Animal Biochromes by Different Authors

Some biochromes are found principally in plants (e.g., flavins) and hence are not relevant to animal signals, whereas others are rarely found in the integument (e.g., blood hemoglobin) and hence manifest coloration in other ways, such as blushing when blood flows close to the skin. Because of molecular variation, densities of pigmentation and anatomical structures, the same group of biochromes may give rise to various colors of the integument (table 4-II , p. 98). From the table it appears that carotenoids, melanins and fuscins provide particularly favorable systems for creating different colors with minimum chemical changes, and hence might be readily changed with minimum genetic change in animal populations. Although I could find no systematic reviews of biochromes used in animal signal-colors, it appears from extensive literature that melanins and carotenoids are the most common bases of coloration in molluscs and bird feathers (also see below).

Melanins are polymers, often attached to a protein as a granule, and catotenoids are long hydrocarbon molecules with conjugated double-bonds and a terminal cyclohexene ring. In birds, melanins are readily synthesized from tyrosine (an amino acid), but carotenoids must be ingested as part of the diet. Carotenes are pure hydrocarbon carotenoids sometimes found in avian soft-parts (legs, beaks, bare skin, eyes), and their oxidation yields xanthophylls, the principal carotenoids of feathers and soft-parts. Ingested carotenoids may be deposited directly or changed chemically before deposition as biochromes.

Inspection of table 4-II (next page) shows that where as long-wavelength reflection may be created by various. biochromes, the violets, blues and greens are rarer. Carotenoid-protein complexes exist as blue biochromes, and blues are often created by schemochromes as well (see next section). The comparative rarity of extractable blue dyes from animals made the Royal Tyrian Purple of the ancients particularly values (extracted from the shell of the gastropod Murex). Even today, blue an purple remain symbols of royalty.

Table 4-II

Biochrome basis of Animal Coloration (compiled from various sources)

Green-reflecting biochromes are so rare (table 4-II) that most animal green is created by schernochrome-biochrome combinations (see next section). Green is the sensation produced by light from the center of the visible spectrum so that a green biochrome must have two absorption bands: one at each end of the spectrum. This fact predisposes green coloration based on biochromes to show dichromatism (ch 3). It may be that blue and green optical signals are relatively rare (especially among birds) because of the difficulty of evolving mechanisms to create these colors, although an obvious problem with green signals is that they would not contrast well with a background of green foliage (see ch 7). Green is a common color among some reptiles and insects, where it presumably functions in concealment (ch 6).

Although the comparative research comprises a very scattered literature, it appears that virtually any class of biochrome may be expected in any major animal phylum. There are, however, certain emphases and restrictions in particular phyla, and especially in taxonomic classes. Of the animals that commonly communicate optically, cephalopods utilize ommochromes in the integument and melanin in their ink; arthropods have a diversity of pigments, including ommochromes and pterins, with porphyrins being rare; and vertebrates use most of the biochromes, especially melanin.

More specifically among arthropods, the crustaceans possess carotenoids, bilins, melanins, ommochromes, pterins, flavins and metallic-based pigments. Virtually every sort of pigment has been discovered in one insect or another. Vertebrate pigmentation is more restrictive, with bony fishes, amphibians and reptiles possessing primarily melanins, carotenoids, pterins and flavins; bird feathers relying primarily on melanins and carotenoids (also structual colors); and mammals being virtually restricted to melanins in hair and skin.

Despite the array of biochromes available to animals for signals, puzzling metabolic constraints on the evolution of such pigments exist. As A. Brush (in lit.) has pointed out, animals synthesize melanins but not carotenoids, and even within a class of animals such as birds, metabolic pathways of biochrome chemistry vary without apparent pattern. "Why aren't alternative pathways used? What determines the degree to which molecules can be modified by animals? Is it a matter of the presence of the proper enzyme array, the energetics of synthesis or the stability of the end product?" Future research in these areas may yield valuable clues to understanding the coloration of animal signals.

A final point concerning biochromes seems particularly important. Because one class of biochrome may provide various colors (table 4-11), very small genetic shifts can provide dramatic color differences among animals. Test (1942) first pointed this out in flickers (Colaptes), the yellow-shafted and red-shafted forms being considered separate species until quite recently. The yellow under-wing and under-tail coloration of the former becomes reddish orange in the latter, and the male's mustache mark is black in the former and red in the latter (there are other subtle differences). It appears that the red-orange-yellow series in flickers might be due simply to concentra- tion of biochrome. Brush (1970) found in a convincing analysis of red and yellow coloration of tanagers (Ramphoeelus) that different concentrations of the same carotenoids were responsible for the whole array of colors. Such results suggest a two-part conclusion: coloration may be evolutionarily labile within a single biochromic system, yet in a more general sense evolutionarily constrained to that system once it is evolved.

Schemochromes

Animal coloration could also be based on any of at least five physical principles explained in ch 3: refraction, Mie scattering, Rayleigh scattering, diffractive interference and thin-layer interference. The spectral dependence of the index of refraction (eq 3.8) may be utilized to select certain colors from complex bioluminescent organs containing transparent lenses, but in general does not appear to be an important factor in animal coloration. Mie scattering by particles that are relatively large compared with the wavelength of light is probably an important factor in diffuse white coloring, but is not known to have noticeable spectral effects on animal coloration. Most schemochromic coloration is therefore due to Rayleigh scattering or interference.

Blue coloration of bird feathers usually appears to be due to schemochromes rather than biochromes, but just why the blue jay, for instance, is blue remains a matter of dispute (see Brush, 1972). The classical explanation is that minute melanin particles in the feathers are responsible for Rayleigh back-scattering of short wavelengths being absorbed by a heavy melanin layer beneath. Dyke's (e.g., 1971) studies of parrot feathers,however, suggest an entirely different mechanism in which hollow keratin cylinders seem to be responsible for thin-layer interference-reinforcement of short wavelengths. Some green feathers, in any case, arise from a blue schemochrome mechanism plus yellow biochrome.

Better understood than feather schemochromes are those in scales of fishes due to layers composed of crystals of guanine and hypoxanthine, two similar nitrogenous compounds. The crystalline layers are aligned in stacks whose spacing determines the spectral band reflected (see fig 3-9, p. 71) according to thin-layer interference. The primarly manifestation of this schemochromic mechanism in fishes is the nearly total reflection of visible light due to the overlapping of reflecting stacks with different spacing, providing most fishes with their mirror-like, silvery reflection.

Iridescence is the name given to metal-like reflectance of a whole range of colors, and is always due to interference. In molluscs, the iridescence in mother-of-pearl is due to calcite layers that create primarily thinlayer interference. It was once thought that non-iridescent blues could be attributed to Rayleigh scattering and iridescent blues to interference, but as pointed out above interference may give rise to non-iridescent colors as well. In sum, Rayleigh scattering (often called Tyndall Scattering in the literature on schemochromes) has been reported for reptiles, birds and mammals, and in the latter is the cause of blue eyes such as mine. Diffractive interference appears to be rarer, but has been reported as the basis of coloration in some aquatic invertebrates and some insects. Thin-layer interference is the process most commonly attributed to schemochromes, being reported for a cephalopod, various crustaceans, transparent insect wings, solid colors of butterfly wings, beetles, fishes, reptiles and birds.

color-change

The rapidity with which animals change color varies, and is governed by various mechanisms. Seasonal changes in birds and mammals are due primarily to molt of the plumage or pelage, respectively, although in some cases it is due to wear of the feathers or hairs. A well-known example of wear is the assumption in spring of the black throat-patch of the male house sparrow as the gray tips of the feathers wear away. Holt is often triggered by changes in day-length and mediated through hormonal control.

More rapid, hormonally controlled color changes are due to edema-like swellings, particularly in the genital areas of many mammals that become pink due to surface vascularization during the reproductive season (Hailman, 1977a: figure 30). In other cases, specific seasonal biochromes are laid down under hormonal control, as in the beak of the starling, which turns from black to yellow in early spring.

The most rapid color changes seem always due to chromatophores: special structures that alter color through dispersion or concentration of biochromes. Most animals that communicate optically have evolved chromatophoric capacities to some extent, including crustaceans and cephalopods (plus a few insects) among invertebrates, and fishes, amphibians and reptiles among vertebrates. The covering of inert feathers and fur make such mechanisms improbable in birds and mammals.

There are three kinds of chromatophores. Cephalopods possess chromatophoric organs with muscle fibers that can act rapidly, the changing color of the Mediterranean octopus having been known to Aristotle. Crustaceans have chromatophoric syncytia that appear to work by streaming action of cytoplasm carrying the biochrome. Vertebrate chromatophores are cellular , and the color-changes of the famous African chameleon were also known to Aristotle.

Chromatophores are frequently named for the biochromes they contain or the coloration they create, so that melanophores contain melanin and xanthophores produce yellow coloration. Guanine-containing chromatophores are guanophores or leucophores, and those responsible for iridescent sheen are iridophores or iridocytes. When the biochrome is concentrated into a tight ball (punctate state) it lends little to the animal's color, but when dispersed (reticulate state) covers any underlying biochrome and hence changes the animal's color. Invertebrate chromatophores may contain two or more pigments, but cellular vertebrate chromatophores each contain a single pigment. Despite their different cellular mechanisms, cephalopods and fishes seem capable of altering their color faster and through a greater range than any other animals.

Types and Origins of Signals

Animals use the mechanisms of light-modulation just reviewed to create spatiotemporal arrays of photons that constitute the signals of optical communication (ch 2), The types of signals and their evolutionary origins are reviewed with examples and illustrations in Hailman (1977a), which is here summarized and extended in certain ways. The detailed aspects of optical signals that actually encode information, which is to say the sign-vehicles (ch 2), must be determined experimentally in each individual case to discover which variables of the signal affect the re- ceiver's behavior.

types of signals

Visual stimuli may be divided roughly into two broad classes: (a) spatially unpatterned light that encodes information primarily by physical intensity and perhaps spectral distribution, and (b) spatially patterned light that gives rise, for example, to the visual images with which we humans are introspectively familiar. Spatially unpatterned ambient illumination is important in the control of such behavior as migratory restlessness in birds, phototactic responses in many animals, circadian rhythms and so on, and may be sensed with simple photoreceptor organs such as ocelli. Spatially patterned light is more important in animal communication, and requires more sophisticated photorecepotrs such as the compound eyes of arthropods or the image-forming, camera-like eyes of vertebrates.

The first distinction in classifying optical signals is whether the source of patterning is the sender itself (intrinsic signals) or some object fashioned by the sender (extrinsic signals). Extrinsic optical signals include such objects as scratch-marks on a tree, the display bowers built by male bowerbirds to attract females (see figure 1 in Hailman, 1977a) and the type-symbols on this page. Wilson (1975; 186) proposes that extrinsic signals be called “sematectonic”.

Intrinsic optical signals are naturally divided into those created by bioluminescenceand those created by reflection of ambient Light , The latter constitute the overwhelming majority of animal signals and have appropriately received the most ethological attention. Reflected signals have both behavioral and morphological elements, and the behavioral elements may be partitioned into three types as shown in fig 4-1. The sender may assume a special orientation with respect to the receiver, may create a special body shape and may conduct specific movement; all the combinations of these behavioral elements are known from animals, as reviewed and illustrated in Hailman (1977a). Accompanying these behavioral elements are morphological elements of signals that may be partitioned into modifications of the reflecting surface itself and larger structural aspects that alter the visible shape of the reflecting surface, such as display plumes, antlers, etc , (see Hailman, 1977a for examples and illustrations)

Fig 4-1. Types of behavioral signals represented as a Venn diagram of three elements of intrinsic_כreflected- light signals.

The variety in signals due to specializations in the reflecting surface are manifest in the spectral reflectivity, spatial pattern of reflectance and type of reflectance (diffuse vs specular; fig 3-5, p. 65). As discussed in the next chapter, receivers analyze spectral reflectivity according to sensations of hue, saturation and brightness, together comprising the perceived color of the object viewed. Because the spectral reflectivity of signals is rarely determined by physical measurement, ethologists tend to describe this variable according to their own perception of color. In many cases it creates little mischief to describe signals in terms of hue, saturation and brightness (ch 5) , but it is important to keep in mind that these are not physical descriptions.

The spatial variables in reflected-light,intrinsic signals are summarized in table 4-III. For a discussion of the specialized topic of bioluminescent signals the reader should consult Lloyd (1977), I have not attempted a classification of variables in extrinsic signals, which are less common than Intrinsic signals in animal communication, but many of the aspects of intrinsic signals given in table 4-III apply to extrinsic signals as well.

Table 4-III

A Classification of Optical Signals (based on Hailman, 1977a)

phylogenetic origins

Animals may cue on any morphology or behavior of other animals and hence such morphology and behavior constitute signals. However, this volume focuses on those aspects of morphology and behavior that have been influenced during the course of evolution by selection pressures to enhance their efficiency as social signals: ritualized signals or displays (fig 2-4). All such ritualized signals are presumed to be evolved from non-signal behavior (Darwin, 1872; Tinbergen, 1951) and therefore have phylogenetic origins (ch 1) that may prove crucial to identify in order to understand signal characteristics. For example, the differences in threat signals of sparrows and gulls can be most parsimoniously explained by differences in the fighting methods from which the signals evolved (Hailman, 1977a: figures 23 and 24).

The behavioral elements of optical signals have been evolved from a huge range of non-signal behavioral patterns (Hailman, 1977a provides examples and illustrations). Virtually any non-signal behavioral pattern may be changed during phylogeny to become a signal, and this includes autonomic responses as well as movements mediated by skeletal musculature (table 4-IV) . In addition, certain opticai signals have been evolved from behavior used to generate signals in other channels, a process I called secondary ritualization (Hailman, 1977a).

The phylogenetic origins of morphological elements are not so well understood, and have been little studied. In the final major section of this chapter, various selection pressures that affect morphological elements, such as surface coloration, are discussed. It seems possible that some coloration was originally evolved for purely physical reasons, such as retardation of integument-wear or protection from damaging ultraviolet radiation (next major section), and other coloration was evolved for visual concealment from predators or prey (ch 6). Then these pre-exisiting capacities were ritualized into social signals.

Table 4-IV

Phylogenetic Origins of Behavioral Elements of Optical Displays, as Reported in the Literature (based on Hailman, 1977 a)

behavioral stereotypy

Behavioral elements of intrinsic optical signals (table 4-III) tend to be highly similar from one performance to the next, except when variability itself is used to encode information. Behavioral stereotypy is not restricted to optical signals, however, so it is instructive to review other causes of stereotypy in orientation, shape and movement.

Specific orientation with respect to social companions may be used for receiving as well as sending signals. Animals with highly encephalized sensory receptors, such as terrestrial vertebrates, often turn their heads toward social companions to optimize sensory information about the appearance, sounds and smells of their conspecifics. In some cases, these simple orientational responses have themselves been the evolutionary origins of optical signales involving orientation, as noted previously.

Strikingly stereotyped postures or body shapes also look like optical signals, evenwhen they are not. For example, male mammals of many species exhibit the characteristic flehmen posture after nuzzling the vaginal region of a female, which posture facilitates the chemical assessment of the female's fluids by a special organ in the roof of the male's mouth. Again, however, in certain species the non-communicative posture has been the phylogenetic origin of an optical signal (Hailman, 1977a). Other stereotyped postures have nothing to do with sensory perception and are not known to be origins of optical signals. For example, cormorants (Phalacrocoracidae) and anhingas (Anhingidae) have feather structures that differ from those of ducks and other waterbirds in that the feathers of the former become waterlogged (Rijke, 1968). Therefore, these birds often seek a perch in air after a prolonged period of diving, spread their wings in a strikingly stereotyped posture, and thereby dry out the feathers (fig 4-2).

Fig 4-2. Wing-drying posture of the anhinga, a dramatic and stereotyped pose that looks like an optical signal. (From a photograph by the author.)

Stereotyped movements are so common and widespread among animals that these provide the greatest potential source of confusion between signals and non-signal behavior. Any time an optimum manner of doing something exists, natural selection may favor a stereotyped movement. Thus animals may move with a certain gait that provides the most efficient locomotion, groom with fixed movements that most efficiently clean the body surface and so on. Stereotyped movements are therefore the least diagnostic of behavioral elements in optical signals.

Finally, it is useful to emphasize again that behavior may appear to be generating optical signals when in fact the communication is in another channel. For exampie, tail-slapping of fish is a stereotyped, oriented movement that communicates by displacement waves in the water (Tinbergen, 1951), and release of communicative chemicals may involve stereotyped postures and movements (Hailman, 1977a: figure 13). Again, such behavior may secondarily give rise to optical signals per se.

biochromes and radiation-absorption

In the analysis of optical signals one is primarily concerned with the spectral reflectivity of animal surfaces (table 4-III). However, biochromes act by absorbing light they do not reflect (eq 3.4) so it is possible that certain biochromes have evolved for their absorption rather than reflection properties. In such cases, the surface coloration of the animal may be a secondary outcome that is largely irrelevant to the primary function of the biochrome. There are two well-known phenomena of this type, and some suggested cases that have not been well studied.

Short-wavelength photons have relatively higher energetic content than those of longer wavelength (eq 3.2), and ultraviolet radiation in particular is highly penetrating in biological tissues. Melanin absorbs radiation across a broad spectrum, and the dark peritoneal linings of lizards and possibly other desert animals are almost certainly evolved for UV-protection of internal organs (Porter, 1967). The extent to which animals have evolved melanic body covering for protection from ultraviolet radiation enjoys no consensus (Hamilton, 1973). It is possible that other biochromes confer similar advantages. Brush (1970) found cartenoid concentration in feathers of Ramphocelus tanagers to increase with altitude, where UV flux becomes greater, although he does not attribute the correlation to UV-shielding. Burtt (1977) found that UV penetrated white feathers most readily, carotenoid-bearing feathers significantly less, and melanin-containing feathers least of all.

There is little doubt that the heat arising from radiation absorbed by melanin at the surface of some animals is an important factor in their energy-balance (Porter et al., 1973). However, there is doubt about how important a factor this is indetermining animal coloration (e.g., Hamilton, 1973). The role of melanin in avian feathers and mammalian fur is particularly complicated, since reasonable but opposite arguments can be marshalled for expected effects. One may argue that black insulation would transfer heat to the animal or that it would capture heat at the periphery and reradiate the energy, thus keeping the animal cooler. Porter(pers. comm.) has recent experimental evidence favoring the latter mechanism in mammalian hair, but there is also evidence for the former action. For example, the Himalayan coat-color pattern in domestic rabbits and Siamese cats consists of melanic fur at the periphery of body projections such as pinnae, nose, tail and paws. These distal tips are subject to cooling to a greater extent than proximal areas of the body. Reared in warm environments, the fur does not deposit melanin, but in cold rearing environments the coloration develops and then remains constant through successive molts. Thus the Himalayan pattern seems adaptive for accumulating heat rather than dissipating it.

Burtt (1977) has recently provided convincing evidence that the nonfeathered legs of wood warblers (Parulidae) are colored in association with temperature extremes: dark-legged species stay north longer in the fall, overwinter farther north and return north earlier in the spring than do light-legged species. There seems little question that melanin and possibly other biochromes are involved in heat-balance of many animals, although just how the mechanisms work may vary among species.

In addition to these two major absorptive functions of biochromes may be added a few other suggestions. Loomis (1967) points out that vitamin-D is synthesized by ultraviolet irradiation, and therefore melanic skin of man might prevent oversynthesis in environments with high radiation loads. Menaker (pers. comm.) suggested that the dark caps of some birds might be effective in regulating the amount of light that reaches photic receptors in the brain, which are responsible for controlling daily and annual cycles of rhythmicity (see Menaker, 1968). Finally, J. Baylis (pers. comm.) points out that the black peritoneal lining of transparent bathypelagic fishes is thought to be an adaptation to keep light in: the melanin prevents the bioluminescence of prey in the fish's gut from being detected by other animals.

biochromes and non-radiative protection

The deposition of biochromes in animals may provide forms of protection that have nothing to do with the ambient radiation on the animal. The primary example of this is the resistance to abrasion conferred by melanin deposition, but other examples have been reported as well.

Dwight (1900) appears to have first suggested that melanin in avian feathers retards wear, but this idea was supported only by arguments and anecdotes until Burtt (1977) recently conducted the first experimental demonstration of the phenomenon. Using a geological fossilcutting gun under controlled conditions, he found that melanin-containing feathers showed much greater resistance to abrasion than did other feathers, whether or not they contained carotenoids. D.L. Fox (1962), however, suggests that carotenoids may also confer some advantage in abrasion-resistance. The mechanism of resistance to abrasion is not fully understood, but is probably not due to meianin per se. Needham (1974: 158) notes that melanin bonding to keratin strengthens the keratin, and melanin-containing feathers often show thickness surrounding the areas of deposition. It seems likely that at least some dark coloration in animals results specifically from selection for melanin-induced resistance to abrasion, as in the black wing-tips of otherwise white seabirds (Averill, 1923). Burtt (1977) has shown the highly abraded areas of wood warblers, such as the dorsum and the flight feathers, are all covered with melanin-bearing feathers, often colored by other biochromes as well.

Needham (1974) cites two other types of protection from biochromes in insects. Quinones tan insect exoskeletons, and it appears that dark Drosophila integument is less wettable than light-colored covering. Furthermore, the offensive odors produced by some arthropods appear to be due to p-benzoquinones from the exoskeletons. These odors, which offer protections from predators, are correlated with the chromatic properties of the molecules, so that coloration may be a by-product of the selection for chemical defenses. Lastly, some insects are known to deposit nitrogenous wastes in the exoskeleton as a method of excretion that prevents water-loss. Because the exoskeleton is molted once or twice a year, this form of protection against dehydration may give rise to particular external coloration as an irrelevant consequence.

In addition to these protective functions, biochromes act to protect animals from predation through concealing coloration and mimicry. For purposes of this volume, such coloration is considered a form of optical communication of "misinformation" and is accorded its own chapter (ch 6).

biochromes, schmochromes and photoreception

Certain biochromes and schemochromes may be evolved to aid visual processes, and these may be considered under two general categories : specializations within the eye itself that have little effect on the surface color of the animal, and specializations of the external covering that are directly related to photoreception.

Three kinds of biochromes and at least one kind of schemochrome are known from eyes of animals. In many invertebrate eyes there are screening pigments that affect the spectrum of light allowed to penetrate to the photoreceptors. Somewhat similarly, oildroplets in the retina and epithelium of vertebrate eyes may contain carotenoids (Hailman, 1976c). The receptors themselves contain visual pigments that change chemically upon absorption of light as the primary process in photoreception (ch 5). Finally, all vertebrate eyes except those of albinos are lined with a pigment epithelium containing melanin, the principal function of which is to capture stray photons that penetrate the retina without being absorbed by visual pigments. Functionally similar pigment is found in many invertebrate eyes. In addition to these biochromes in the eye, the pigment epithelium or retina itself in vertebrate eyes may contain guanine plates similar to those in fish scales that reflect light. Such tapeta lucida and analogous structures function by thin-layer interference, and refleet photons to provide a second chance for their capture by visual pigments during very low levels of illumination. In high light levels the tapedial plates may be shielded by extensions of the pigment epithelium. Tapedal reflection of light may be seen as eyeshine in nocturnal animals, but eyeshine seems an artifact resulting from man's creation of highly directional light sources, and it is unlikely that eyeshine is a factor in animal communication under natural conditions.

At least three photoreceptively related functions of biochromes on the body surface have been proposed. Needham (1974: 163-177) cites carotenoids and possibly porphyrins as bases of dermal light sensitivity, and states (p. 177) that "probably in all animals there is a general dermal light perception, whether or not discrete eyes also are present." Two possible uses of biochromes to aid conventional photoreception concern coloration near the eyes. Ficken (Ficken and Wilmot, 1968; Ficken et al., 1971) has suggested that lines projecting anteriorly from the eyes are used by birds and other animals to sight prey for a feeding-strike. Burtt (1977) points out that such lines are often dark and may instead be antiglare adaptations, although the two functions are not mutually exelusive. The antiglare strategy is also used to aid human vision, as when football players put lampblack on their cheeks to diminish specular reflectance or when airlines paint the metal below airplace cockpit windows flat black. It seems likely that black around the eyes could serve to diminish reflection in animals other than birds (e.g., the black mask of the raccoon), but a similar pattern of coloration would also be predicted as a deceptive mechanism for hiding the eyes (ch 6).

These various uses of biochromes and schemochromes for other than signaling functions are summarized in table 4-V. It is unlikely that this list is complete, and the table serves primarily as a reminder that signal coloration of senders may often be an evolutionary compromise with other advantages conferred by biochromes.

Table 4-V

Νon-signal Uses of Biochromes and Schemochromes Established or Suggested