The compound eyes of trilobites are different from the true compound eyes of today in that their lenses were made of the mineral calcite. Calcite is widespread on Earth - chalk is calcite, but granular, so that it appears white via scattering. Scattering causes structural colour - a white or blue appearance depending on the size of the scattering elements. The elements, or granules, are relatively large in chalk, which causes all wavelengths in white light to be reflected equally and in all directions. And, as Newton demonstrated, when all wavelengths combine the light appears white. But if the calcite is formed slowly, a perfect crystal results, completely free of granules. This type of calcite is crystal clear, and was the ingredient of trilobite lenses. Today, calcite lenses are found only in bristlestars, relatives of starfish. And these lenses are not part of an eye as such, rather a component in a composite light perceiver comparable to that of some bristle worms. Although all relied on calcite lenses, there were two distinct types of compound eyes in trilobites - holochroal and schizochroal.
Schizochroal eyes were big, but not because of the number of facets they contained, which were surprisingly sparse. Instead they owed their size to the vastness of each facet - up to a whole millimetre in diameter, a dimension not even approached in today’s compound eyes. A boundary region separated each facet from its neighbours, and the lenses were either elongated prisms or came in two parts that locked together, one above the other.
The two-part lens is interesting. In addition to flying a kite into a thunderstorm in an attempt to understand electricity, the American diplomat and scientist Benjamin Franklin was also famous for inventing bifocal glasses in the eighteenth century. These offered the wearer the choice of seeing objects either near or far with accuracy. The trilobite schizochroal eyes with the two-part lenses did the same thing, so that they could see both tiny prey within grasping range and enemies approaching from a safe distance. The graded material lenses first identified by Maxwell in his breakfast herring are also to be found in some schizochroal eyes. Making a direct comparison, these trilobite eyes may have worked like those superposition eyes with graded lenses, such as are found in a group of crustaceans called mysids. On top of this, a new type of eye was found recently in a living insect called a strepsipteran, and this may unlock the secrets of the enormousness of the schizochroal facets.
Figure 7.13 Photographs of holochroal (above) and schizochroal (below) trilobite eyes.
Strepsipterans are tiny insects that parasitise wasps. The eye of the male strepsipteran has only fifty lenses, as opposed to seven hundred in the similar sized Drosophila fruit flies, but they are each relatively huge. Each lens is linked to its own retina, and the nerves serving the individual retinas cross over, so that a complete picture can be assembled in the brain with everything in the right place. This form of imaging, which was probably echoed in the schizochroal trilobite eye, falls somewhere between that of compound and simple eyes.
Efficient as it may seem, and strepsipterans aside, the schizochroal eye was confined to just one group of trilobites known as the Phacopina. Phacopina lived up until 370 million years ago, but they did not evolve until the very end of the Cambrian period, around 510 million years ago. The schizochroal eye evolved from the trilobite eye of most interest in this chapter - the holochroal type. This has a significantly earlier origin.
Holochroal eyes generally contained more facets than schizochroal eyes, where each facet was relatively small. The lenses were simple - thin and biconvex (‘oval’), like the lenses of magnifying glasses. They were packed together in a square or hexagonal formation, where neighbouring lenses were touching. But exactly how the holochroal eye functioned is something of a mystery. The real problem is that part of the eye may or may not be missing in the fossils. We do not know. The calcite lenses have preserved well thanks to their chemistry. But were there further focusing elements lying behind these that just haven’t preserved because of an unfavourable chemistry for fossilisation? In one way, taking a strict view of their position in the eye, the calcite lenses of trilobites could be more comparable to the thick corneas of modern compound eyes. In which case we would expect a further focusing element, or lens, to be lying just beneath. But then again, maybe the calcite lenses were the only focusing elements of holochroal eyes in trilobites, and maybe these were quite adequate.
So the internal architecture of the trilobite eye, as it is known, cannot provide the information needed to make comparisons with modern compound eyes. But there are some clues to be found on the outside. Those holochroal eyes with square-shaped facets may be comparable to the reflecting type of superposition compound eye, where the facets are square for a reason. These are the facets lined with mirrors, and the mirrors perform the focusing. And then those holochroal eyes with hexagonal-shaped facets may have worked like today’s apposition compound eyes, with stark similarities on the outside. If these inferences are correct, we could predict the environments or lifestyles of the trilobites.
One shrimp today changes its eye throughout its development from juvenile to adult. As a juvenile it has an eye with hexagonal facets - an adaptation to its bright, shallow-water environment. This apposition eye is good at producing sharp images, but not so good at collecting all the light available. Fortunately there is a plentiful supply of light for this juvenile. But as it grows, it migrates to deeper waters, where light becomes more limited. So the apposition eye is shed during the moult to adulthood, and is replaced by a superposition eye with square, mirror-box lenses. This new eye has quite the opposite properties; although not effective at forming sharp images, it can make the most of the light available. All of this evidence suggests that trilobites with hexagonal-shaped facets lived in shallow waters, and those with square-shaped facets lived deeper, or were active at night.
On the other hand, square and hexagonal shaped corneas may be consequences of lens-packing geometry alone - the way they are squashed together. Because an alternative exists, unless completely preserved holochroal-type eyes are found in the future we will never know with certainty exactly how this organ worked and, consequently, how its hosts viewed the world. For the purposes of this book, it is enough to say with confidence that trilobite eyes produced visual images - trilobites with these eyes could see. Now we should move on to more important matters.
Although the origin of the holochroal eye now appears to pose a big question, it is a question that has never been properly addressed. Without the line of enquiry maintained in this chapter, there is little justification for pursuing such an otherwise unimportant goal within the gravity of science. But in this chapter we have been deliberately building up to the very first eye in existence on Earth. Through a process of elimination we have arrived at the holochroal compound eye of trilobites. And from here we are in the hands of palaeontologists. We must rely on the fossils to help provide us with a date for that very first holochroal eye - the very first eye. The fossils do not let us down.
The oldest trilobites known are from the Lower Cambrian - the earliest part of the Cambrian. So far, so good. But we can be even more precise than that. The very first trilobites evolved at the very beginning of the Cambrian, around 543 million years ago - and they were equipped with holochroal compound eyes. Before this date there were neither trilobites nor eyes on Earth. So it is worth a look at those first trilobites and their eyes.
The oldest known, well-preserved trilobite eyes were described by Euan Clarkson, an expert on trilobite eyes at the University of Edinburgh, and his colleague Zhang Xi-guang from the Chengdu Institute of Geology in China. Working on material from south central China, they found particularly interesting compound eyes in two species of trilobite - Neocobboldia chinlinica and Shizhudiscus longquanensis.
Xi-guang and Clarkson used acid to dissolve the limestone slabs excavated from their Lower Cambrian site. The trilobites were freed from their matrix and were ready to be studied in electron microscopes. They had been particularly well preserved, thanks to the protection afforded by a phosph
ate coat, and so fine details of their optics could be uncovered.
Neocobboldia bore a thick lens in each facet of its eye, and the lens was free from spherical aberration, the problem created when rays entering different parts of the lens are focused on to different planes. But it showed no signs of a graded material lens like that of the herring or some compound eyes today. How was spherical aberration avoided? The answer lay in a sophisticated design, involving a precisely curved divide within the lens. This ‘intralensar bowl’ design was not new to science - Huygens and Descartes had invented something similar in the seventeenth century - but here trilobites were proving that it really worked.
The lenses are less well preserved in the eyes of Shizhudiscus, although they are obviously simpler in design - they are biconvex (‘oval’). All things considered, these eyes conform to the rules of the holochroal type and so they could produce visual images in the very Early Cambrian. And nice panoramic pictures, too - images of anything positioned on the trilobites’ horizon.
Figure 7.14 The intralensar bowl design in the lenses of some trilobites; light rays striking all parts of the lens are focused in the same plane. An identically shaped lens without the intralensar bowl is shown for comparison.
In fact there are a number of trilobite species known from the very beginning of the Cambrian, and all have eyes, although they are not so well preserved. Even Charles Doolittle Walcott suspected this trend in 1910. And in 1957 the trilobite specialist Frank Raw, of Birmingham University in England, declared, ‘How ancient already in the Lower Cambrian must the compound eye have been.’ This statement was seconded by Euan Clarkson in 1973. Another case of trilobite eyes is that of Fallotaspis from Morocco, around 540 million years old. Fallotaspis had large eyes. The list continues . . .
Figure 7.15 Time ranges of genera within the seven families of trilobite, showing the occurrence of different kinds of eye (after Euan Clarkson, 1973). Note that the very first trilobites, living at the base of the Cambrian, bore (holochroal) eyes.
To reiterate, there is one curious but common fact emerging from the trilobite eye data. Many species of trilobites with eyes came into existence around 543 million years ago . . . but not a single species before that time. The trilobites without eyes entered history a little later, in geological time. So 543 million years ago the Earth witnessed the first trilobite . . . and the first eye. Five hundred and forty-three million could be a magic number.
An important question in this study is, ‘How quickly can an eye evolve from its forebears?’ Fossil evidence implies that eyes existed 543 million years ago but not before. Not, say, 544 million years ago. But surely an eye cannot simply evolve overnight? Surely it has to pass through a sequence of intermediate stages, probably within intermediate species in the evolutionary tree? These intermediate species must have fallen between the completely eyeless ancestor and the first eyed ascendants. If some of these intermediate species could see, albeit in a rudimentary way, and if they existed millions of years before the first ideal eye, then maybe animals acquired vision earlier than 543 million years ago. Maybe the introduction of eyes on Earth was staggered; perhaps sharp images were seen millions of years after blurred images, with visual precision increasing systematically. Dan-Eric Nilsson and his colleague at Lund University, Susanne Pelger, have deciphered a series of intermediate stages for the evolution of a camera-type eye. More than that, they have calculated the time needed to complete the sequence via the process of evolution. This is just the data we need.
At the beginning of this chapter we examined light perceivers that could detect light levels but not form visual images. They were not eyes. But some light perceivers were more efficient than others, and it is likely that a more efficient type originated from a less efficient type during its evolution. To make their predictions, Nilsson and Pelger applied this logic.
A patch of light-sensitive skin was used as a starting point. This dents inwards, and becomes increasingly infolded to form detectors escalating in their sensitivity to the direction of light. This assumption is quite acceptable since all the intermediate stages can be found functioning in animals today. It is important that each link in the chain can exist in its own right. The opposite of this was once used to criticise evolution, and even clouded the thoughts of Darwin himself, as suggested in the epigraph at the beginning of this chapter. To justify this further, we can explain why all animals don’t possess the theoretical ultimate eye. The intermediate stages, or conceptually substandard visual organs, do exist today because their host animals cannot handle the information loads supplied by the next conceivable stage on the road to a fully formed camera-type eye - Darwin had no reason to be concerned. Back on the evolutionary road, we have reached a ‘cup eye’ that cannot form proper images. We have also reached a junction. Close the entrance to the cup even more and we have the pinhole eye of nautilus. Then again, begin to grow a lens and another path has been taken - the path to the camera-type eye typical of vertebrates.
Figure 7.16 Nilsson and Pelger’s predicted evolution of a camera-type eye, like that of a fish. The sequence begins with a flat patch of light-sensitive cells sandwiched between a transparent protective layer and a layer of dark pigment. A graded-index lens appears at stage 6. Reproduced from a 1994 paper by Nilsson and Pelger with permission from the authors.
Nilsson and Pelger were more than realistic in assuming that a light receptor will change by just 1 per cent of its length, width or protein density during each evolutionary step in the eye direction. But even with such a pessimistic approach, the whole sequence from light-sensitive patch to the eye of a fish would require only two thousand of these tiny modifications in sequence. That may not seem enough, but as Michael Land and Dan-Eric Nilsson point out, if two thousand sequential modifications of 1 per cent are applied to the length of a finger, then it becomes long enough to bridge the Atlantic Ocean.
We know that proteins need not evolve from their chemical beginnings. A study of flatworms revealed that similar proteins exist in the eyespots (not true eyes) and touch/chemical detectors. In the eyespots, these are the proteins that react to light, comparable to those in the retina of an eye. So a head start may be gained towards eye evolution by borrowing the proteins of other detectors.
Now for the calculation of time needed for these modifications to take place, which is really what we are interested in. Again, caution was the name of the game when Nilsson and Pelger made their assumption about the slowest rate of evolution - a 0.005 per cent modification from one generation to the next. In reality, the rate would probably be faster. For instance, the light receptor pigments of modern crustaceans show an evolution that is considerably more rapid than expected. And verily the word ‘pessimistic’ entered the title of Nilsson and Pelger’s original paper, which made their result seem even more remarkable. They found that the eye of a fish could evolve from its rudimentary beginnings in less than 400,000 generations. Assuming each generation is completed within a year, this result suggests that an efficient, image-forming eye can evolve in less than half a million years. Now that really is a blink of an eye on the geological timescale.
This is a camera-type eye and we have established that the first eye was compound. But in their definitive book on the optics of animal eyes, called Animal Eyes, Michael Land and Dan-Eric Nilsson were beginning to picture the evolutionary sequence of the compound eye. They claimed that arthropods ‘probably originated from a worm-like ancestor that already possessed a rudimentary compound eye - possibly a loose collection of eyespots’. Independently, the Australian biologist Richard Smith mapped the changes needed to form the compound ‘eye’ of a bristle worm. A loose collection of eyespots also appeared in Smith’s sequence. And the number of links expected in the chain leading to a fully functioning eye was on a par with those in Nilsson and Pelger’s predictions for the camera-type eye.
Like the proteins of the retina, other parts of the body involved in the process of light perception seem quite accommodating
to these calculations on the eyes themselves. Nilsson and Pelger’s time prediction would be meaningless if development of the visual processing centre in the brain was lagging behind that of the eye. In 1959 the biologist von Bekesy demonstrated that the effects caused by sound can be mimicked by vibrating the skin. This demonstrated that the ear and skin shared certain common features, namely nerves, in the processing of sensory information. But what does this mean for the evolution of the eye? Well, it is conceivable that nerves used by one sense can be ‘upgraded’ for use by two senses. And if the senses of hearing and touch can share features, then so might vision and touch. In this way, the nerves needed to service an eye would not have to evolve from a vestigial beginning - they would have a head start. Then there is a possible helping hand in the brain department. Parts of the brain, it appears, may be capable of converting from touch to vision. Dan-Eric Nilsson suggested that the compound ‘eye’ of ark clams and bristle worms evolved from chemical detectors that were inhibited by light. So the evolution of the eye itself appears to be the limiting factor, or at the back of the pack, on the evolutionary road to vision - the remainder of the system can simply be adopted. Indeed, there were other sense organs surrounding the eyes of trilobites, and the original light perceivers may have borrowed nerves from these.
Now we can calm our own nerves that may have been jangling while we gave the compound eye just one million years to evolve - at least if it was to fit with our fossil evidence. It seems that our demand has been met - one million years is plenty of time for an eye to evolve. Now we can paint a picture of 544 million years ago, where light sensitive patches were evident in the ancestors of the Cambrian trilobites. Then we can paint another picture of 543 million years ago, just the other side of the Cambrian border, where a trilobite proudly flaunts its eyes. Between the two pictures the light-sensitive patch had evolved into an eye.
In The Blink Of An Eye Page 25
