Fish vs. Cephalopods

In marine environments, cephalopods clearly rank at the top of the Protostomes, in terms of size, intellegence, and trophic dominance. At the top of the Deuterostomes are either the modern teleost fish or the marine mammals. For my purposes, I am going to use fish as the Deuterostome example, because they reflect a more primary adaptation to aquatic habitats, and occur in a wider range of sizes and habitats. (Competition between fish and marine mammals may be relevant to the more recent success of the teleosts, compared to other groups of fish, but that's a story for another day.)
There are many similarities between cephalopods and fish that suggest that they are a prime example of convergent evolution driven by their competition for the same resources, in the same recent era. This view has been forcefully stated in the review article:
"Cephalopods and Fish: the limits of convergence", A. Packard, Biological Reviews 47: 241-307 (1972). This article discusses "the remarkable fact that cephalopods are like fish in almost every feature except their basic anatomical plan."

A contrasting view has been presented in:
"The constraints on cephalopods: why squid aren't fish", R. K. O'Dor & D. M.Webber, Canadian Journal of Zoology, 64: 1591-1605 (1986). They summarize by saying: "Despite the functional similarities created by competition in the nekton, ... squid are no longer so much competing with fish as trying to stay out of their way."

1. Cephalopods are all marine. Fish live in both marine and freshwater environments, and there is in fact strong evidence that all marine fish have descended from ancestors living in fresh waters.

2. Cephalopods and fish compete over the same range of sizes. Approximately 0.01 to 10 meters body length.

3. Both are active, mobile, predators. All cephalopods are carnivorous, but fish cover a wider range of diets and include filter-feeders.

4. Each of the major forms of modern cephalopods has a fish counterpart:
In tide pools and rocky substrata: Octopus vs. tide-pool gobies and moray eels.
On sandy bottoms: The cuttlefish, Sepia , vs. flatfish such as plaice, sole, halibut. Sit on the bottom inconspicuously until food gets close enough to catch.
Fast, pelagic, predators: small squids (Loligo ) vs. herring, etc.

5. Both squids and fish have similar adaptations for rapid swimming:
Streamlined bodies
Fins -- provide lift to balance gravitational sinking
-- can locomote by waves propagated along the fins
Schooling to increase swimming efficiency and protection(?).
[Has been observed that a lone squid will sometimes join a school of similar- sized fish.]
Fish have also discovered the energetic advantages of a burst and coast mode of swimming.

Flying fish and squid have both discovered advantages of this form of burst locomotion. [up to 15 mph and 50 m leaps.] Squids use water jet for additional accleration after they leave the water.

However, the major means of propulsion is different. Although some fish can rapidly close their opercula (gill covers) to produce a jet of water and accelerate the body, the major form of fish locomotion is undulations of the body or tail fin. In squids, jet propulsion is the major form of locomotion. In cuttlefish, both jet propulsion and fin waves are important. In octopus, jet propulsion is especially used for escape responses, while crawling over the rocks with the tentacles is probably the most common form of locomotion.
The squid body is highly adapted for rapid jet propulsion, with most of the body weight in the mantle muscles, and with giant nerve fibers to ensure rapid, synchronous contraction. Can produce mantle pressures as high as 0.4 atmospheres and produce accelerations of up to 3.3 g.
Nevertheless, D'Orr and Webber conclude that "the best performing squid studied to date still uses more than twice as much energy to travel half as fast as an average fish." Fish like tuna, optimized for propulsion by the tailfin, probably do even better.
Why? The answer comes from the physics of propulsion, rather than from biology. For large swimming animals, operating at high Reynolds Number, the forward momentum of the fish or the squid is maintained by imparting equal backwards momentum to the water. For the squid, this momentum is initially imparted to the water leaving the mantle cavity, at high velocity. For a fish, this momentum is initially given to a much larger volume of water, at a lower velocity.
For equal amounts of backward momentum, the larger volume of water will have lower kinetic energy, because its velocity is lower and kinetic energy is proportionalto v2. Therefore, the least power is used by accelerating a large volume of water to a low velocity. It's similar to the reason why a heavy gun is more powerful than a light gun. That is, more of the energy is transferred to the bullet instead of to recoil of the gun.
Further study shows that the relative inefficiency of squid propulsion gets worse as the animals get larger. For small, juvenile squids, the difference is less. It has been suggested that this explains why cephalopods grow rapidly, reproduce, and die (semelparous), in contrast to fish that grow and reproduce over many seasons (iteroparous).

6. The metabolism of squids is based on rapid digestion of high protein diets, the use of the amino acids for growth (mostly muscle), and when necessary the degradation of muscle and other protein to supply energy. There is very little energy storage as lipid (max. 6% of body weight vs. 15 to 25% in fish) or glycogen (less than 0.5% of body weight).

7. Both fish and squids have evolved adaptations to control buoyancy, but the mechanisms are very different.
In fish, bones increase weight, but stored fat decreases it. The fish swim bladder increases buoyancy, but is also controllable to permit adjustment to neutral buoyancy at different depths. The fish swim bladder is inherently unstable, because if the fish descends to a greater depth, the pressure increases, the air in the bladder is compressed, and the density of the fish increases, tending to cause sinking. This is overcome by active secretion of oxygen into the swim bladder, by mechanisms that are capable of generating the high gas pressures needed at great depths.

In squids, there is little fat storage, but density is sometimes decreased by altering the ionic composition of the body fluids -- replacing Na with NH4, etc. Greater changes are brought about by air compartments -- in the shell of the older groups - ammonites, Nautilu s -- or in the cuttlebone of squids. The mechanism involves active transport of solutes out of the compartments, so that there is osmotic withdrawal of water, and air accumulates in the spaces instead. This cannot produce a pressure to prevent collapse, so the collapse of the compartments must be prevented by the rigidity of the walls, which means added weight. This mechanism only works down to depths of about 240 meters. The limit is set by the osmotic pressure of the body fluids, which are osmotically similar to sea water. Below that depth, squids will be heavier than water, and must swim to avoid sinking. However, this mechanism is inherently stable.

8. The high level of activity of cephalopods requires a high rate of oxygen delivery. One advantage of jet propulsion is that the water flow over the gills is automatically increased as the animal swims faster. In fact, there is so much water flow over the gills that only a small portion of the oxygen, as little as 8%, is actually used. Cephalopods are handicapped by the fact that oxygen transport, in the blood in combination with Hemocyanin, has much less capacity --about one-third that of a typical fish. This is primarily due to the relatively larger protein mass required to bind each oxygen molecule (compared to hemoglobin), but it is not clear why this hasn't been improved upon. The low capacity is partly compensated by the fact that more of the capacity can be utilized, since the oxygen partial pressure in the gills is always high. The net result of this is that the squid must pump 4x the blood flow to swim half as fast as a typical fish.

9. Perhaps because of some of these basic physiological limitations, cephalopods have survived by evolving their behavioral capabilities. A striking convergence is the evolution of sensory organs -- balancing organs and eyes -- that have amazing resemblances to the vertebrate organs.

10. Eyes. Cephalopods and vertebrates have simple eyes like a camera, with one lens projecting an image on the retina, in contrast to the compound eyes of arthropods, with a lens on each retinal element (ommatidium).
There are other animals with simple eyes, but they all have relatively small numbers of retinal cells (<10000) and are primarily photodetectors rather than image interpreters. Cephalopods and Vertebrates have more than a million retinal cells per eye. Vertebrates have more, especially in the fovea, and therefore better acuity (human about 30x better than octopus).
It's easy enough to multiply retinal cells to increase the number, but it is not worthwhile unless:
a -- the optical system is good enough to form a sharp image on the retina.
b -- the brain can interpret the information from large numbers of retinal cells.

There are some major differences:
a) Cephalopods have one type of retinal photoreceptor, in direct position. Fish have several types (rods and cones), in inverted position.
b) Cephalopods have a nerve going from each retinal photoreceptor to the brain, with relatively little interaction (information processing) at the retinal level. In fish, a significant amount of information processing occurs in the retina, and there are fewer axons in the optic nerve than there are photoreceptors.
c) Fish rely on retinal pigment migrations to adapt to different light intensities without using an iris. Cephalopods have an iris - an irregular horizontal slit - and pigment migrations are slow.

These differences make the basic optical similarity even more remarkable:
Both have a spherical lens, concentric with nearly spherical retinal surface.
Focal lengths are 2.1 to 2.6.
aperture about f 0.8.
A homogeneous spherical lens in water can have this focal length only if the refractive index is 1.67 = glass. But maximum known refractive index of biological materials is about 1.58 (in the center of these eye lenses).
A homogenous glass lens with r=1.67 has terrible spherical aberation. The image is useful only at very low aperture. Fish and squid lenses can be removed and examined, and are found to produce very good images.
The only known solution to this is to have the refractive index vary with radius. The optimal curve for refractive index vs. radius can be calculated, but it is difficult to get good measurements on real lenses, and so there is argument about exactly how the problem is solved. In any event, it is solved, and in a dynamic manner that allows the eyes to grow (at least 10x in fish) while still having the proper refractive index distribution. [The eyes of giant squid grow to 40 cm diameter!]
The lens is made of a central core of dead cells, covered with layers of live epithelial cells. These cells have very high concentrations of proteins called crystallins. It has recently been discovered that the crystallins are not unique proteins, but are proteins that carry out other functions in other cells at lower concentrations. Most vertebrate lenses contain several crystallins, and some crystallins have turned out to be known enzymes found in other cells. Both squids and vertebrates make use of enzymes as crystallins, and in one case, a glutathion S-transferase, is used by both squids and mammals. (For more information, see Tomarev & Piatigorsky, Eur. J. Biochem. 235: 449-465 (1996).) In some cases, there is solid evidence that the same gene is used to produce the enzyme in some cells, and the high concentrations of protein that function optically in other cells. The use of several different crystallins may be important in producing a refractive index gradient.

11. Chromatophores. Both cephalopods and fish make extensive use of chromatophores to control their appearance. In fish, this is mostly done for camoflage. In octopus, there are additional usages including displays that serve for sexual identification. In both cases, chromatophores work by dispersing or concentrating pigments, with the pigment affecting the appearance more when it is dispersed. However, the mechanisms are quite different. Vertebrate chromatophores involve intracellular movements of pigment granules along microtubules, and are typically regulated by hormones. Cephalopod chromatophores are operated by muscles that expand or contract the pigment sacks, and are regulated by the nervous system, so that responses can be very rapid.

12. What can cephalopods see? Most of the work on this question has been done with Octopus . or with the cuttlefish, Sepia . The first requirement is to find a way to get the octopus to tell you what it sees. To do this, experimenters have taken advantage of the learning abilities of Octopus . The question is rephrased to: Can the octopus tell the difference between two different visual stimuli? This is tested by trying to train the octopus to give different behavioral responses to the two stimuli. For example, we can ask whether the octopus can tell the difference between _ and | by presenting these stimuli to an octopus in an aquarium. If the octopus comes out and examines the _ it is rewarded with a piece of food, and if it comes out and examines the | it is given an electric shock. It will soon learn to respond positively to the _ stimulus and not respond to the | stimulus, so we have proven that it can see the difference between these two stimuli. These tests also provide information about mental processes, in addition to visual capabilities. For instance, after training as just described, using white shapes, tests show that the animal can generalize this discrimination to black shapes, and to shapes that are 4 times larger.

Other training sessions show that Octopus can see the difference between objects of different size (2x), independently of distance; black vs. white; and the direction of polarized light: _ vs. |, or \ vs. /. However, it cannot discriminate \ vs. / shapes, and has a difficult time learning to discriminate between an upright square and a square that is rotated 45 degrees.
The _ vs | shape discrimination becomes difficult or impossible after removal of the balancing organs (statocysts). The reason for this is that information from the statocysts is used to control the muscles that rotate the eye, so that the pupil slit is horizontal. The animals without statocysts will learn the _vs. | discrimination if the stimuli are carefully oriented with respect to the pupil slits. This shows up a major difference between information processing in vertebrates and cephalopods. In a vertebrate, information from different sensory modalities, such as inner ear and eyes, is integrated in the brain.This is not done in cephalopods for statocyst information and visual
information, or for tactile information and visual information, as far as has been found. Some other limitations on the intelligence of these animals have been found:
An octopus in an aquarium will never learn to go around a glass plate to reach a crab on the other side. It just tries to go through the plate, at least until it reaches the plate and spreads out over it. On the other hand, if an octopus in an aquarium is given a pile of rocks, it will try to crawl under them and hide, rearranging them if necessary. It will just as happily hide under a glass plate.
This is just a sampling of the experimental approaches to trying to understand the neural capabilities of Octopus.

One question that this raises: Given the relatively advanced neural capabilities of Octopus , and the excellent vision, why hasn't it advanced further? In particular, there is no evidence of tool usage. Perhaps this is because it is difficult to know where a long flimsy tentacle is, in contrast to the situation with rigidly jointed appendages, and the inability to use visual information in combination with proprioceptive information from the tentacles, to achieve fine motor control.

For more information, I recommend the book "Octopus" by Martin Wells (Chapman & Hall, 1978)