Universal rhodopsin

As it happens, a number of other proteins can act as light receptors, and they too provide some illuminating cases of convergence. Yet the rhodopsin molecule seems to offer various advantages: as Russell Fernald156 has remarked, in the context of vision although an opsin 'is not the only way to detect light... it has proven irresistible for use in eyes'.157 Rhodopsin is remarkably widespread, and apart from its near-universal adoption in the various visual systems this molecule has long been known to occur in the bacteria. These organisms exist as two kingdoms, the Archaea and the Eubacteria, of which a typical representative is E. coli. The most notable examples of rhodopsin are in those Archaea referred to as halophiles. As their name indicates, they live in salt-ponds, where their immense numbers frequently colour the lagoons red or purple, which is the colour of the pigment associated with their rhodopsin. The molecular architecture of this so-called bacteriorhodopsin,158 with seven a-helices spanning a lipid membrane, is the same as the molecule in our retina. So, too, is its basic dependence on light, but in the bacterial variety the energy of the Sun's photons is used to drive a proton pump159 to transfer hydrogen ions (i.e. protons, H+) as the basis of the cell's energy system, specifically by the synthesis of ATP.

Bacteriorhodopsin has received a vast amount of attention. This is principally because of its now-classic role as an ion pump and the desire to elucidate its detailed structure as a protein.160 Yet a remarkable fact and an intriguing inference concerning bacteriorhodopsin have emerged. The former concerns its widespread distribution. Its occurrence in archaeal bacteria is well known, but only recently has it been appreciated that in oceanic ecosystems this molecule occurs in the eubacteria (in which organisms it is referred to as proteorhodopsin), where it appears to play a key role in the harvesting of the sunlight in a way analogous to photosynthesis.161 Nor does the distribution of bacteriorhodopsin stop there. As we have seen, various single-celled protistans, including some green algae, have eye-spots. It is hardly surprising, therefore, to find rhodopsin in the single-celled Chlamydomonas and colonial Volvox.162 This story has had its convolutions, especially with the identification of a membrane-bound protein, referred to as chlamyrhodopsin, in the eye-spot. This, however, has a molecular structure that is markedly different from that of genuine rhodopsin.163 It is now clear that not only does rhodopsin itself occur, but it evidently has a bacteriorhodopsin-like structure.164 Bacteriorhodopsin, however, is even more widespread, and it also occurs in the fungi. One instance is in the parasitic chytrid-iomycete fungi. Their reproductive cells, known as zoospores, are equipped with a rhodopsin-bearing photoreceptor that guides this swimming stage towards the light.165 In many, and perhaps all, these fungi the rhodopsin employed is equivalent to the bacteriorhodopsin. The intriguing question, therefore, is whether this type of rhodopsin is strictly the same as that found in our eyes, or is in fact convergent. The latter seems to be more likely. First, it has long been appreciated that there is no discernible similarity in the amino acid sequences when the two types of rhodopsin are compared.166 Furthermore, despite the overall similarities of molecular architecture there are some specific differences.167 In their review of rhodopsins, John Spudich and colleagues168 conclude, the genome sequence data presently available and the three-dimensional structures of the molecules themselves argue that nature discovered retinal [i.e. the vitamin A derivative] twice, and both times found it useful, when solvated with seven helices, for photosensory signalling as well as other phototransduction functions. The two-progenitor hypothesis would require that archaeal sensory rhodopsins and mammalian rhodopsins have converged on remarkably similar mechanisms of receptor photoactivation. Such closely similar mechanisms could result from 'likely reinvention' determined by the inherent properties of retinal as a chromophore.169

The remarkable similarity of the molecular architecture of the two rhodopsins may, therefore, well be convergent. If this is the case it is perhaps not surprising that bacteriorhodopsin is also very sensitive to substitutions at key sites. Thus, in a way analogous to substitution at key sites controlling colour vision, so meddling with site 85, for example, can completely destroy the ability of the bacteriorhodopsin to pump protons.170 And the parallels may be even closer, because it appears that some types of bacterial rhodopsin are also capable of a colour discrimination,171 suggesting once again the possibility of a universal property. Wherever life sees light it will be through the agency of a rhodopsin, and where it sees red it will depend on key amino acid substitutions. So, too, the sentient inhabitants of Threga IX will also enjoy those spectacular red sunsets, by camera-eyes, through transparent crystallins, and by means of rhodopsin.

Light reception by organisms is therefore very widespread. Rhodopsin may have evolved twice, but other proteins can also act as light-receptors. Opsins may well be best, but the other examples are also instructive in the context of both light sensitivity and also convergence. For us, the role of light is effectively synonymous with vision, yet, as is widely appreciated, there are also perceived cycles of light and dark. Most notable, perhaps, are the circadian rhythms of the 24-hour solar cycles, of day and night.172 Also important is the longer-term seasonality, with the most familiar effects in higher latitudes of autumnal colours in plants, as well as hibernation or migration in animals. Such rhythms and changes are very widespread and are monitored by light-sensitive molecules. Of central importance are the so-called cryptochromes, cryptic neither in being difficult to find nor in establishing a photoreceptor function, but because until quite recently their true structure was problematic. Cryptochromes are proteins with a flavin component (and linked to a vitamin B2-based pigment) that are sensitive to blue light.173 They occur, for instance, in the retinas of mammal eyes, where they play an important role in the setting of the circadian clock, effectively the discrimination between night and day.174 The importance of this clock is familiar from the disorientating effects of jet lag; more seriously, disruption of the circadian clock is implicated in major malfunctions during the night shifts of nuclear reactors and chemical plants.175 Danger might also lie elsewhere. For example, disruption of circadian rhythms by the widespread use of artificial light may also be implicated in the increased prevalence of breast cancers.176

Given their role in light sensitivity, it is unsurprising to learn that cryptochromes occur also in algae177 and plants.178 What is more interesting is that the plant cryptochromes appear to be very ancient, whereas those of animals are more recent and an independent invention.179 Such a convergence is less surprising in one sense, because in both cases the precursors of these cryptochromes lie with a group of proteins known as the photolyases whose function is to repair DNA damaged by light, especially by ultraviolet radiation.180 Curiously enough, although humans have cryptochromes most probably we lack photolyases, thereby increasing our vulnerability to ultraviolet radiation and, by implication, skin cancers.181

smelling convergence

Given that key molecules required for vision, such as rhodopsin and the crystallin proteins, evolved in single-celled organisms this suggests that given time182 and the adaptive value of light discrimination then the evolution of the eye seems to be a near inevitability. And so fascinating is the story of the evolution of the eye, not to mention our particular dependence on vision, that we tend to forget other types of sensory world, even when walking the dog at night with the air full of bats: to us other senses, such as those of olfaction and echolocation, are almost closed books. Yet the principle of evolutionary convergences should encourage us to think, not only of recurrent re-emergences of given senses, but also about deeper similarities of sensory input, such as those of hearing, generation of electrical fields, and perhaps even the sense of smell. In this last case there have been some curious experiments, such as those by the remarkable Victorian polymath Francis Galton. Today he is probably best remembered for his interest in hereditary principles and the related application of eugenics,183 the monstrous nature of which was articulated with characteristic prescience by G. K. Chesterton.184 Galton's autobiography185 is still entertaining, even if it reveals an individual to whom self-doubt was an almost complete stranger.186 Perhaps that may explain some of the less appealing aspects of Galton. Thus, in his absorbing account of the recognition of the importance of fingerprints in forensic investigation, Colin Beavan187 argues that for all his contributions to this area, driven in part by the hope that fingerprints would yield eugenic insights, Galton was less than fair to the other pioneers, not least to one Henry Faulds. Indeed, Beavan shows that behind the facade something unpleasant, but not so rare, lurked: 'Other people's success aroused a venomous jealousy in Galton.'188 Well, well.

In his memoirs Galton mentions a paper he had written, entitled 'Arithmetic by smell'.189 Galton's interest was evidently sparked by what is called non-verbal representation, crudely whether we (and by implication other species) can think and handle abstract concepts without recourse to words. So Galton decided to investigate one of the further reaches of mental representation. Here is what he wrote:

Leaving aside Colour, Touch and Taste, I determined to try Smells. The scents chiefly used were peppermint, camphor, carbolic acid, ammonia, and aniseed. Each scent was poured profusely on cotton wool loosely packed in a brass tube, with a nozzle at one end ... A squeeze of the tube caused a whiff of scented air to pass through the nozzle. When the squeeze was relaxed, fresh air was sucked in and became scented by the way. I taught myself to associate two whiffs of peppermint with one of camphor, three of peppermint with one of carbolic acid, and so on. Next, I practised small sums in addition with the scents themselves, afterwards with the mere imagination of them. I banished without difficulty all visual and auditory associations, and finally succeeded perfectly. Thus I fully convinced myself of the possibility of doing sums in simple addition with considerable speed and accuracy, solely by imagined scents. I did not care to give further time to this, as I only wanted to prove a possibility, but did make a few experiments with Taste, that promised equally well, using salt, sugar, quinine, and citric acid.190

It is easy to smile at Galton's equivalent to an olfactory abacus,191 yet these convergences of sensory input may have more subtle evolutionary implications. Consider, for example, the star-nosed mole (Fig. 7.7).

figure 7.7 The star-nose mole, a typical mole in terms of its powerful fore-limbs, but remarkable for its tentacular nose, with its nostrils clearly visible. (Photograph courtesy of Professor K. C. Catania (Vanderbilt University, Tennessee).)

We have already met the moles as a splendid example of fossorial convergence among the burrowing mammals. The star-nosed variety is unusual in constructing its burrows in wet and boggy ground, and it also spends a considerable amount of time swimming.192 This mole inhabits the eastern part of North America, but it may be a recent immigrant because fossils have been found in Poland.193 It is almost blind, but by way of compensation has an extraordinary nose consisting of 22 mobile and fleshy appendages, somewhat like a star (Fig. 7.8). This nose, however, is not used for smelling, nor do the tentacles serve to capture food directly. Rather, as Ken Catania has so elegantly shown, this remarkable structure retains its sensory powers for the detection of prey, but is modified in a most interesting fashion.194 The surface of the nose is densely studded with button-like sensory structures, known as Eimer's organ (Fig. 7.8).195 It is estimated that the nose carries about 25 000 of these mechanoreceptors. Not surprisingly, the nose is exceedingly sensitive, and although it is only about a centimetre across it is supplied with five times as many nerves as run into the human hand. But the real surprise is that not only is this nose ultra-sensitive, but more importantly it shows some remarkable parallels to the eye. In particular, one region of this nose towards the

figure 7.8 Details of the nose of the star-nosed mole. Above, detail of a tentacle, studded with the sensory Eimer organs. Opposite, electron micrograph of the nose. The analogue of the fovea is represented by the two central lobes beneath the nostrils. (Photographs courtesy of Professor K. C. Catania (Vanderbilt University, Tennessee).)

figure 7.8 (cont.)

central area is especially well supplied with nerve endings. As such it can be compared directly to that region of the eye known as the fovea, where the retina possesses exceptional sensitivity. Both this part of the mole's nose and the fovea show a comparable degree of nervous acuity.196 In making comparisons between nose and eye Catania is unequivocal about their similarities when he writes, 'The many surprising parallels between the somatosensory system of the mole, and the visual systems of other mammals, suggest a convergent and perhaps common organization for highly developed sensory systems.'197 Interestingly, in another underground mammal, the naked mole-rat, which we met earlier (Chapter 6, p. 142) in the context of the convergent evolution of eusociality, nearly a third of that part of the brain allocated to touch perception (the somatosensory cortex) is rededicated to the prominent incisor teeth.198 Effectively these teeth take the place of the usual rat whiskers, and given that the naked mole-rat is for all intents and purposes blind, it is not surprising that the area of the brain normally employed in vision appears to have been almost entirely taken over by the somatosensory cortex. So, in a way analogous to that of the star-nosed mole, these naked mole-rats probably 'see' with their teeth.

'Seeing' with a nose (or even teeth) allows for some intriguing speculations. Imagine a planet ejected early in its history from its parent solar system, which as we saw earlier may be an all-too-common occurrence (Chapter 5, p. 86). A position whirling through the Stygian gloom of interstellar space might seem to make such a lonely world an improbable abode for life,199 especially at an advanced level. Probably so, but let us imagine also an abundance of radioactive elements in the planetary crust, whose continuing breakdown by decay serves to maintain subterranean ecosystems at viable temperatures. Perhaps in caverns, which on Earth extend at least 1.5 km beneath the surface, any animals would be 'seeing' in much the same way as the terrestrial star-nosed moles and naked mole rats. Some hint of how adept such hypothetical animals might be also comes from studies of how the blind mole-rats navigate their complex burrow systems in a way that would do justice to a rat in a maze.200 Sensory cues may possibly come from the Earth's magnetic field, but it also seems conceivable that the mole-rat is able to build a cognitive map: one more hint of mentality in 'dumb' animals (see Chapter 8).201 There are even instances at the surface of the Earth in which eyes are lost, or more strictly eaten, without deleterious effects. One particular example concerns a population of tiger snakes living on Carnac Island, off the western coast of Australia. Silver gulls attack these snakes and remove the eyes, but the snakes are in a sense 'over-designed' and they are able to survive202 using other sensory mechanisms.203 It is also worth noting that while snakes cannot re-grow eyes, where regeneration is possible, as in arthropods, amputation of a sensory appendage may lead to its replacement by other appendages with their own neurological (and sensory) characteristics.204

The remarkable proboscis of the star-nosed mole provides a number of other useful and relevant insights into the tension between novelty and constraint in evolution. In one sense the origins of this mechanosensory nose are not so peculiar; the related talpids, that is the other true moles, also have highly sensory noses, again studded with Eimer's organs. More specifically, comparisons with moles such as Townsend's mole (a species of Scapanus) give a good idea of how the star-nose might have arisen.205 Even so, in its evolved form the star very much represents a biological novelty. Many animals have fleshy appendages of various kinds, which routinely employ a particular developmental pathway that depends on specific developmental genes (notably one called distal-less). In the star-nosed mole, however, the appendages develop in a unique manner, without parallel elsewhere.206 Why this mole should pursue such a novel developmental route when the genetic template for appendage formation is readily available, and indeed presumably coded for in the limbs of the mole, may seem rather odd. There are, however, particular constraints. Catania and his colleagues suggest that the need to keep the neural connections, without which of course the nose is functionless, means that the standard method of 'let's build an appendage' simply is not available. Finally, remembering the topic of convergence, it is worth mentioning that the development of the elaborate nose is not without consequences for the architecture of the skull. In particular, the chewing apparatus is quite modified (e.g. weaker dentition and more fragile mandible), and as such is convergent on other animals that are either vermivores or termite eaters and thus tend to swallow their prey.207

As noted above, the star-nosed mole does not employ its spectacular proboscis for olfaction. But here, too, those animals that do have a sense of smell show some striking similarities. So it is that despite major anatomical differences the basic mechanism of olfaction seen in the insects (usually located in the third segment of each antenna) and vertebrates (the nose, of course) depends on nervous structures known as the glomeruli.208 These act to connect and integrate the olfactory messages that are then conducted as electrical signals via the nervous system to the brain, where the discrimination of odours takes place. To be sure, the proteins (and thereby genes) that serve to bind the odours, as the first step in olfaction, are different,209 but as John Hildebrand and his colleagues have emphasized, despite the molecular genetics being quite different, our sense of smell and that of the insects (and other arthropods) works on the same principles,210 and so encompasses such specific examples as the sex pheromones.211 Not only do the properties of the insect glomeruli 'have precise counterparts in vertebrate olfactory bulbs',212 but Nick Strausfeld and John Hildebrand go on to remark, 'Basic types of olfactory information coding are remarkably similar in both vertebrates and insects' with various excitatory responses to odours strongly indicating 'common strategies for processing odor information at the level of glomerular relay neurons'. These writers are fully aware of the implications when they continue, 'The primary olfactory centers in the brains of vertebrates and insects thus appear to share canonical cell arrangements and common physiological properties ... Furthermore, vertebrates and insects show remarkable parallels in events underlying the development of glomeruli ... Do such commonalities suggest that the glomerulus, as an olfactory functional unit... originated in a common bilaterian ancestor [i.e. to insects and vertebrates]? Or, is common design and physiology the consequence of convergent evolution? Is there a uniquely logical response to common selective pressures such that to construct a glomerulus requires the same set of rules?'213

A definitive answer is not yet possible, but given the differences in the molecular genetics of olfactory receptor coding, the evidence points strongly to convergence. This is because the predecessors of the vertebrates, the amphioxus animal (see Chapter 1), and the insects, that is the aquatic crustaceans,214 do not possess such glomeruli, which indicates that the method of effectively identical olfaction arose independently.

Once again it seems that the options are limited. There may be only a few ways to smell, and, as we have seen in the star-nosed mole, if a nose is not used for olfaction it can end up as the sensory equivalent of an eye. And again there may be deeper similarities pointing towards the inevitability of evolutionary end results. The processing of the olfactory signals, and especially the recognition of particular odours, involves a synchronized neuronal activity that leads to an oscillatory firing of the neurons. The associated networks are not only similar in both the insects and the vertebrates, but they in turn may also find parallels to what occurs in the processing of visual stimuli in the mammals, at least.215 Further evidence for more profound similarities between the processes of olfaction and vision also comes from an investigation of proteins known as arrestins which, as their name suggests, are involved with the termination of a physiological process, such as the quenching of phototransduction in rhodopsin.216 Given that the proteins involved with olfaction have a basic similarity to such visual proteins as rhodopsin,217 it would not in principle be surprising to find arrestins also involved with the olfactory transduction pathway. What, therefore, is the more remarkable is evidence for a specific family of arrestins that is involved in both visual (eye) and olfactory (antenna) transduction in insects.218 This is not the only example of such overlap in molecular transduction in the two systems,219 and it is further evidence for a deeper, and largely unappreciated, commonality of these sensory systems.

the echo of convergence

The richness and versatility of the sensory worlds does not end here. Many animals, in addition to the star-nosed mole, inhabit gloomy and crepuscular worlds, but generate elaborate and sophisticated fields of perception that arguably are as refined and sensitive as any visual or olfactory system. Most familiar in this regard, perhaps, are the extraordinary powers of echolocation possessed by the bats,220 a group that offers other insights into convergence.221 The basic principles of echolocation are fairly self-evident, and are of course employed in a somewhat similar way by many people who are blind.222 So, too, with cats.223 In the case of the fast-moving bats, however, what is particularly remarkable is that even when the animal is very close to its insect prey, where the neural responses can no longer keep track of the returning echoes, the bat still keeps a lock on its prey. Evidently the bat must employ a filtering mechanism of some sort, but how this signal is actually interpreted by the brain is still mysterious.224 Another intriguing feature is that not only will the bat navigate through crowded vegetation in pursuit of its prey, but the hunted moths have acute hearing and are capable of evasive action.225 More extraordinary are those moths that also produce ultrasonic clicks, the purpose of which appears to be to jam the bat's sonar.226

Echolocation is not confined to the nocturnal sonar of the bats. It is also well known in various marine mammals, such as the dolphin,227 but in the context of considering convergence we shall postpone a visit to these wonderful animals because of their involvement in the yet more intriguing area of intelligence (Chapter 10). Concerning the techniques of echolocation there is, however, another fascinating example of convergence that is perhaps rather less known. This involves those birds that inhabit the deep, dark recesses of caves, where they often share their domiciles with the bats. Notable in this regard are the South American oilbirds and the Asian swiftlets, whose saliva-bound nests are eagerly sought for those who find bird's nest soup exceptionally tasty. As their name suggests, the oilbirds have extensive fat deposits, which if rendered provide an oil of exceptional quality and purity. These birds attracted the notice of the great traveller and naturalist Alexander von Humboldt, and subsequently were engagingly described by Donald Griffin.228 Oilbirds are nocturnal. They are also noisy animals, but can they echolocate when travelling to or from their perches in the caves? Griffin listened to them as they poured out of their cave at twilight, remarking, 'Perhaps the clicks were call notes, perhaps they were something analogous to profanity, or perhaps they were symptoms of some other avian emotion whose nature we could not guess.'229 His subsequent experiments, however, showed that the ability to echolocate was genuine, with the series of sharp clicks clearly audible to humans. Continuing work230 has shown that the resolution of oilbirds' echolocation is rather crude, at least when it comes to avoiding discs deliberately suspended in their flight path.

Interestingly, despite having well-developed eyes the oilbirds appear to lack binocular vision,231 which is otherwise a general characteristic of birds and convergent on other vertebrates (p. 162). In some ways this makes the behaviour of the Asian swiftlets all the more extraordinary, since their echolocatory abilities enable them to avoid much smaller objects.232 Although their echolocation does not rival that of bats,233 nevertheless in their natural habitat the pitch-dark caves are filled with flying swiftlets, seldom if ever colliding and readily finding their respective nests. Bats, dolphins, and even some birds, therefore, have all entered a sensory world that is apparently almost completely unknown to us (note 227). It is fascinating to speculate what sort of echolocatory image these animals 'see'. Although it is now time to turn to the even stranger world of electrical perception, it is worth remembering that as the commonalities of convergence continue to emerge so we may find that apparently alien perceptions, and even mentalities, may not be as remote as has sometimes been imagined. Even to enter the mind of a bat may not be as difficult as is sometimes supposed. We return to this topic in Chapter 9 (pp. 265-266).

shocking convergence

Of the sensory systems it is perhaps those involving the generation of electrical fields that are the most remote from human experience. Perhaps it is not surprising that it has been suggested that the exquisitely sensitive nose of the star-nosed mole possesses an electrical sensitivity, but this proposal is viewed with considerable caution.234 Even so, such sensitivity is certainly possessed by some other mammals, notably the primitive monotremes.235 It is among the fish, however, that we find not only some of the most remarkable examples of electric generation (and perception), but also - and by this stage are we really surprised? - yet more striking examples of convergence. Thus, electrogeneration has evolved independently at least six times and in each case entails the modification of muscle cells, albeit from a variety of locations.236 The phenomenon of electrical generation by fish must have been shockingly apparent to incautious waders and fishermen for a very long time, and certainly at least as far back as classical times.237 A famous physician, close to the Imperial family of the first century ad, one Scribonius Largus, wrote of the use of the electrical discharges of the torpedo ray, a familiar denizen of the Mediterranean and Atlantic coasts (although it is found in all oceans), for the relief of intractable headaches.238 Such a therapy was also endorsed by Dioscorides, a celebrated medical authority active at about the same time as Scribonius Largus, who, despite a reputation for travelling, spent at least some time in Tarsus.239 He also knew of this treatment, but more alarmingly employed the torpedo's electrical capacity for the treatment of the prolapsed anus.240 Interestingly, the greatest of the ancient physicians, Galen, disputed the efficacy of this galvanic treatment not only for headaches but also for more fundamental matters.241 Nor was familiarity with the electrical properties of certain fish restricted to the classical world. African tribes, for example, were also familiar with the therapeutic powers of such fish, and according to reports provided by Jesuit missionaries and other travellers the inhabitants of Ethiopia employed the discharges for a variety of purposes, ranging from the expulsion of evil spirits to the control of fevers.

The properties of electrical fish also attracted the attention of both doctors and scientists in nineteenth-century Europe. Among the former were those intrigued by the surge of enthusiasm associated with mesmerism and other examples of what was termed 'animal magnetism'.242 Of the scientists, the most famous was the physicist Michael Faraday, who undertook an extensive series of experiments in the late 1830s. These, not surprisingly, focused on the measurements of the electrical current but also included observations on the killing of prey and its ingestion by the fish. Faraday was a thoroughly hands-on, or in this case, hands-in, scientist. In one such experiment he and two colleagues had their hands in the tank when 'suddenly [the fish] gave a shock which startled us all and was perfectly satisfactory as to

figure 7.9 A mormyrid, or elephant-nose, fish (Campylomormyrus). (Photograph courtesy of Professor C. D. Hopkins (Cornell University).)

the generality of the discharge. Mr Gassiot evidently felt it least', but Faraday dryly continues, '... I daresay Mr Bradley [the] most.'243

The fish Faraday employed was an electric eel from South America,244 which belongs to a freshwater group known as the gym-notids. They show a series of quite remarkable convergences with an effectively unrelated group, the mormyrid fish (Fig. 7.9),245 which inhabits the lakes and rivers of Africa.246 In both gymnotids and mormyrids the twin processes of electrogeneration and electrorecep-tion are strongly convergent, although it should be noted that the underlying neurological mechanisms are certainly different.247 These fish have been the subject of intensive experimentation, and the functional significance of their electric signalling and responsiveness has given glimpses into an extraordinary electrical world, which might, however, have analogies with our more familiar sensory modalities. Experiments typically involve the recording of electrical activity and observation of behaviour, but while an important strand of investigation concerns studying how individuals react, it is also possible to 'fool' the fish with an artificial equivalent, consisting of an electric dipole connected to a computer that can respond to the electrical signalling of the fish.248

Almost without exception, the electric organ is derived from muscle, not from nervous tissue,249 but unsurprisingly the cells (or electrocytes) are specialized and commonly form a structure equivalent to a stack of gelatinous discs with a rather remarkable series of interpenetrating stalks.250 As might also be expected, the electric organ itself has evolved. Primitive and advanced arrangements can thus be recognized, but there is evidence in this organ for both reversion as well as convergence.251 Control of the electric organ is from the brain, via particular nerves known as the electromotor nerve axons. The resultant electric organ discharge (or EOD) that is transmitted into the water varies remarkably in duration, frequency, number of peaks, and polarity according to the species concerned.252 There is, moreover, one important difference between the gymnotids and mormyrids inasmuch as with one exception the mormyrids produce the electrical signal as discrete pulses, whereas the gymnotids produce an effectively continuous signal as a wave form. In the case of the gym-notids there is also some evidence that the signals associated with communication, especially important in the sexual context, may be generated in a different region of the fish from those concerned with the imaging of the surrounding environment.253

An interesting possibility is that one of the driving forces towards an increasing complexity of the signals may have arisen as a defence against those predators that are capable of detecting the electrical impulses. Thus, by shifting the frequency of part of the signal the fish become more difficult to detect,254 although it certainly does not render them immune to attack.255 But this is not the end of the matter, because the enhanced electrical signal can then be utilized in sexual communication.256 The gymnotid and mormyrid fish therefore live in what is effectively an electrical world in which the signals are received by the electroreceptors and then transmitted to the brain, which allows the animal to perceive inanimate objects,257 especially when encountered in a novel context,258 and to recognize or communicate with other inhabitants of the murk, be they friends or foe. It is also interesting that the two types of electroreceptor found in the skin, to detect direct and alternating currents respectively, are again very similar in both gymnotids and mormyrids.259

Because the environment in which these fish swim is effectively one of electrical uproar arising from the competing signals of other fish, it is hardly surprising that each species produces a highly specific electric signal. Such call signals, analogous to those issued by individual aeroplanes, are doubly important because the general electrical racket produced by dozens, if not hundreds, of electric fish nearby is further augmented by the frequent tropical thunderstorms with the associated lightning, adding to the 'electrical cacophony [and]... background roar'.260 Not only is there distinction between species, but in some cases either sex or the juveniles of a particular species also produce a characteristic and distinct waveform. For the most part, however, the signals of individuals of a given species appear to be identical. This may, however, be a simplification because in one species of gym-notid the signatures of individuals are discernibly different, a factor of considerable importance in social contexts such as defence of home territory.261 But there are exceptions, effectively a convergence, where up to four species produce a similar electrical signal. It comes as little surprise to learn that in such cases the fish aggregate in schools.262

In this electrical Babel separation of signals is, therefore, essential. What happens, however, if two fish of the same species produce their respective signals simultaneously? Potentially the result would be problematic, for the two signals would lead to destructive interference, and the whole point of signalling would be lost. To circumvent this problem the fish have evolved a mechanism that is referred to as the jamming avoidance response (JAR) in which the fish changes its signal frequency.263 Sensible enough, but two things are remarkable about this avoidance of jamming. First, the fish are astonishingly sensitive to potential interference. They will respond to signal modulations with an amplitude difference of 0.1% and a timing disparity of 400 nanoseconds or less, each fish shifting its signal pattern in a few microseconds.264 Second, and even more remarkably, the algorithm used by the gymnotids and mormyrids to shift the signal has, of course, evolved independently but it is identical. Nor do the convergences end there, because computationally similar neural algorithms also occur in the owl, where acuity of sensory perception is acoustic rather than electric.265 In the owl,266 not surprisingly, neither the neural mechanisms nor the implementations are identical to those of the electric fish, but this does not prevent the emergence of deeper similarities. So when we look at the general arrangement in gymnotids and mormyrids, it is clear that even though different parts of their respective brains are employed for the interpretation of the electrical signals there are also a number of significant similarities in the neural circuitry.

Given the degree of convergence between the gymnotids and mormyrids267 in terms of electrical activity it comes as less of a surprise that they occupy similar environments, living in waters that are either muddy or otherwise opaque 'blackwater', and typically showing nocturnal activity.268 In such a setting the powers of vision are decidedly restricted269 and, at least so far as the mormyrids are concerned, an early investigator, M. R. McEwan, remarked, that 'The sense of sight of these fishes is not at all keen.'270 This observation was consistent with her study of the mormyrid retina, which demonstrated a number of peculiar structures that probably represent adaptations to the low levels of light.271 Aspects of their life cycles and reproduction are also strikingly similar, including the embryological stage.272 Although the body form of each group is rather different, in terms of their ecology, and especially feeding, there are important parallels in these fish. Thus there is a recurrent tendency to evolve elongate and tubular mouths (Fig. 7.9) with weak jaws and a feeble dentition. In arriving at this arrangement, these fish have opened a new resource in the form of a rich bottom-dwelling fauna of worms and larvae that more orthodox arrangements of mouth and jaw would find difficult to exploit.273 Moreover, in at least some instances the prey is detected by the generation of the electrical field.274 Indeed, Tyson Roberts identified an even closer link between feeding and electrical production when he wrote, 'I would go even further, and suggest that the interrelation between electrical faculties and feeding played a decisive role in the initial divergence of the gymnotoids and mormyroids from non-electrically specialized ancestors,'275 as they entered the twilight zone of the Congo and Orinoco. Nor should this mud-grubbing life lead us to underestimate the extraordinary nature of these electric fish. As already indicated, the jamming avoidance mechanism is clearly a complex response involving a sophisticated set of adaptations. By now it should be clear that these fish occupy a rich world of electrical signals and social communication and, relative to at least some other fish, one of enhanced perception. This is consistent with a neural plasticity that in filtering out the predictable indicates that these fish occupy a contextural environment in which learning and memory are the norm.276

So it appears that, in a way analogous to the star-nosed mole, these fascinating fish 'see', not tactilely but in an electrical world.277 So, too, it seems reasonable to think of other electrically sensitive animals, such as the sharks, also constructing an electrosen-sory landscape.278 Nor need this parallel, however unfamiliar it is to humans, surprise us, because when we look at the brain activity

Was this article helpful?

0 0

Post a comment