Curiosities of evolution - Part I: Birds and viviparity

Evolution by natural selection is often incorrectly thought of as an unbounded path to novelty. I was reminded, when writing about coelacanths and manta rays, that there are some traits that are curiously absent from some lineages. In particular, the birth of live offspring (viviparity) has independently evolved many times in a diverse range of taxa, but not in birds or, probably, any other dinosaur (birds are theropod dinosaurs). Many adaptive explanations have been proposed to explain why this might be the case.

There are many ways that live birth can occur and many ways that the mother can provide nourishment to her offspring. The extant coelacanths, for instance, are what's known as 'oviviviparous', because females retain their eggs within their body until their offspring hatch, at which point the mother gives birth. Manta rays too, are oviviviparous, but mothers supply extensive additional nutrition after their offspring hatch.

In oviviviparous animals, the yolk allocated to the egg can be the only nourishment that offspring receive. But, there is a continuum from total reliance on the yolk, through various amounts of additional nutrition supplied by the mother (matrotrophy), to examples where all nutrition is supplied by the mother without yolk. Matrotrophy can be achieved with or without a placenta, and sometimes in alien ways, such as the cannibalism of siblings and unfertilised eggs within the uterus. Animals that are considered 'truly' viviparous are often only those where matrotrophy is complete.

After the ray-finned fishes, birds are the most species-rich lineage of vertebrates, yet not one species produces live young. The available evidence suggests that all dinosaurs were also egg laying. This is curious because all other major vertebrate groups, with the exception of the agnatha, have multiple examples (at least 120) of the independent evolution of live birth. And, unlike the agnatha, birds have several of the traits thought to be required for viviparity, like internal fertilisation.

Many hypotheses have been proposed to explain the lack of viviparity among birds; flight and the hard-shelled egg are the two explanations that seem to best fit the data. Pterosaurs, like birds, flew and seem to have all been egg laying. And bird's eggs are relatively impermeable to gas exchange when compared to those lineages of reptiles that have evolved viviparity. Indeed, turtles and crocodilians have similar hard-shelled eggs and have also not evolved viviparity.

However, these two hypotheses aren't completely satisfying. The 'flight-as-a-constraint' hypothesis is contradicted by live birth in bats and doesn't explain why none of the many flightless birds and dinosaurs evolved viviparity. The 'egg-as-a-constraint' hypothesis hinges on oxygen being a limitation and the assumption that the eggshell cannot be eliminated through evolution. But birds lay their eggs well before oxygen limitation would become a problem for embryos. And the evolution of viviparity in other lineages seems to begin with a lengthening of the period that eggs are retained before they are laid, only later does the eggshell thin to increase oxygen availability.

Some authors have suggested that viviparity has not evolved because directional selection for egg retention must be absent in birds, or that there is selection against it. The argument behind this hypothesis is that many of the advantages that viviparity may confer are solved by other means in birds and viviparity may have negative effects on mothers. For instance, retaining eggs internally may reduce the predation risk for offspring by decreasing the amount of time they're highly vulnerable to predators. But conversely, it may increase the risk of predation for mothers because they must carry an extra weight that could reduce their ability to escape.

Here too, counter examples can be found. For instance, there may be a selective advantage of egg retention in some species where eggs are a small fraction of maternal body mass and external incubation is costly. Many seabirds travel long distances between nesting sites and feeding grounds and they are also often large relative to the size of the egg. This suggests that carrying the extra weight of an egg is a minor cost and that there may be an energetic advantage to reducing the period of external incubation.

Other authors have suggested that the reason egg retention is short in birds is that their body temperature is too high for developing embryos. The resting temperature of birds is nearly universally 40 - 41 °C, while the optimal incubation temperatures are between 34  °C and 38 °C. Body temperature may, therefore, be a physiological constraint on egg retention in birds. But, correlation is not causation. Optimal incubation temperatures may have evolved because they're the temperatures that eggs are normally incubated at, not because they're the maximum temperatures that embryos can tolerate.

To me, the body temperature hypothesis probably carries the most weight as an adaptive explanation, but it remains untested. And, although the lack of viviparity in birds cries out for an adaptive explanation, we should not assume that the presence of absence of particular traits is the result of adaptive mechanisms. Nevertheless, the lack of viviparity in birds and other dinosaurs is an interesting and unresolved evolutionary question.

References:

Blackburn, D. G., & Evans, H. E. (1986). Why are there no viviparous birds? The American Naturalist, 128 (2), 165-190 DOI: 10.1086/284552 

Anderson, D. J., Stoyan, N. C., & Ricklefs, R. E. (1987). Why are there no viviparous birds? A comment. The American Naturalist, 130 (6), 941-947 DOI: 10.1086/284757

The evolution of living fossils

The term 'living fossil' is a problematic one because its meaning is so frequently misunderstood. The greatest misunderstanding is that a living fossil species has not evolved for tens or even hundreds of millions of years (e.g. anatomically modern coelacanths are know from 409 mya). But, this is completely and utterly wrong. Living fossils are species that are related to, and superficially resemble, other species from the fossil record. But, they can be morphologically distinguished from the fossil species and have almost certainly evolved in ways that don't preserve in stone (e.g. behaviourally, physiologically and immunologically).

Three species of coelacanth. The top two are the fossil species Coelacanthus and Macropoma, while the bottom is the extant specie Latimeria. Note that they are all similar, but easily distinguished.

So, it's a little disappointing when a usually very good science news website perpetuates this misunderstanding by starting a popular science article like this: 

The morphology of coelacanths has not fundamentally changed since the Devonian age, that is, for about 400 million years. Nevertheless, these animals known as living fossils are able to genetically adapt to their environment.

There is nothing at all surprising about populations of living fossils containing enough genetic diversity to adapt to the environment. And there is simply no good reason to assume that their genetic diversity will be any different to any other extant species. 

The paper itself does not make this mistake. The interesting thing about studying the genetic diversity of coelacanths is not because they are living fossils, but because they are considered rare and endangered. A good understanding of genetic diversity within populations and an understanding of gene flow among populations can be very informative for the development of conservation management strategies. And this is the aim of the paper.

The authors obtained genetic material from 71 adult coelacanths from 6 locations across the entire known range of Latimeria chalumnae, the East African coelacanth (L. menadoensis is a second species found off Indonesia). The genetic diversity among the coelacanths was low, as would be expected from their small population sizes. The largest population of 300 - 400 individuals occurs off the Comoros Islands and all other populations appear to derive from it. The greatest genetic diversity, however, was found in populations from Tanzania.

The sites that genetic samples were collected from. The size of the circles indicates the sample size (key at bottom) and the colours indicate the genetic types found in a population (taken from the paper).

The genetic differentiation between the Comoros population and populations in other locations shows that adaptation is still occurring, but that there is unlikely to be much gene flow among populations. Interestingly, and completely unexpectedly, there appears to be two genetically distinct populations occurring at the same locations within the Comoros Islands. It's unclear what factors are driving the differentiation in the two Comoran subpopulations.

The genetic differentiation between the populations along the African coast isn't strong. This suggests that either the populations diverged relatively recently, or that they're evolving slowly. Curiously, the authors argue that their results confirm that the coelacanths are evolving slowly. But, their data can't separate these two hypotheses. Other studies show that the genes which control morphology are evolving slowly (surprise!), but other gene regions are within the evolutionary rates for vertebrates and consistent with rates in sharks, which have similar life histories. It seems more likely, therefore, that the populations have diverged relatively recently, suggesting that populations could be being recolonised from the Comoros Islands after local extinctions. 

So, coelacanths, like other living fossils, are evolving just fine. Their populations off East Africa look to have reasonably good genetic diversity for their population numbers. But, populations will probably need to be managed separately because there is little gene flow among them.

Reference:

Lampert KP, Fricke H, Hissmann K, Schauer J, Blassmann K, Ngatunga BP, & Schartl M (2012). Population divergence in East African coelacanths. Current biology, 22 (11) DOI: 10.1016/j.cub.2012.04.053

High impact science

The exoskeletons of arthropods allow them to do many seemingly preternatural things. Storing energy to generate the fastest limb movements in the animal kingdom is one of those things. Stomatopods, or mantis shrimp, aren't the fastest, but the hero of this story (Odontodactylus scyllarus) can accelerate its killing arms incredibly fast (65 - 104 km s^-2) reaching top speeds of 23 meters per second. To my knowledge, that's the second fastest limb movement ever recorded. And it's performed in water, which strongly limits speed relative to air.

The stomatopod Odontodactylus scyllarus, or peacock mantis shrimp (photo Wikipedia)

Stomatopods are coarsely divided into two groups based on the shape of their killing arms. Some stomatopods have arms that look similar to those of praying mantises, which is where their common name is derived. These stomatopods are called "spearers", as they use their arms to impale soft-bodied prey. The other group, which use club-like arms to break open the shells of other crustaceans and mollusks, are called "smashers". But, there is a great diversity in the form and function of stomatopod arms. The largest stomatopods tend to be spearers, but the fastest are smashers.

A sample of the diverse shapes of stomatopod killing arms (photo Thomas Claverie)

Our speedy pal, O. scyllarus, is a relatively large smasher, growing to 18 centimeters. Although water does hamper the speed of its strike, it also provides an interesting advantage. The fast movement of the club-like killing arm causes a cavitation bubble to form between the club and the point of impact. Cavitation bubbles form in areas where water pressure drops so low it forces a phase change from liquid to vapour. The higher pressure in the surrounding water then causes the vapour bubble to collapse rapidly, releasing a shock wave of sound and a burst of light. The collapse of the cavitation bubble is so violent that it can impart a force as great as the club strike on the prey item.

 

The stomatopod O. scyllarus in action

The double strike of the club and cavitation bubble causes an impressive amount of damage to the hard-shelled prey of O. scyllarus. But, the impact of the strike and cavitation bubble also cause damage to the club itself. The only time this damage can be repaired is when the stomatopod moults. Although O. scyllarus does moult relatively frequently, its clubs are still resilient enough to deliver thousands of blows between moults.

The resilience of the club's hitting surface is due to a complex, three-region architecture that allows small cracks to form, but prevents them becoming large enough to be a problem. The outer 'impact region', which forms the hitting surface, contains a highly crystalised form of the mineral hydroxyapatite. The inner 'periodic region' also contains hydroxyapatite, but here it occurs in the amorphous mineral phase and is interspersed with the carbohydrate chitin, which is arranged in helical stacks. The 'striated region', which forms the back and sides of the club, is made of chitin arranged in circumfrential bands.

Cross sections of an O. scyllarus club, illustrating the three-region architecture. The crystalised hydroxyapatite impact region, the amorphous hydroxyapatite and chitin periodic region and the chintinous striated region. The lower panel shows the distribution withing the club; blue for the impact region, red and yellow for the periodic region and green for the striated. The orange indicates another segment of the killing arm that sits behind the club and acts as the 'handle' (image modified from Weaver et al. 2012).

The periodic region is the 'shock-absorber' of the system and the majority of cracks that form in the club occur in this region. The cracks that form are forced to twist inwards by the helically arranged chitin and are largely prevented from spreading between layers or into the impact region. The striated region of chitin reduces the deformation of the club during strikes and helps to keep the cracks contained within the periodic region. This unique architecture allows the club to survive the damaging forces of many high speed strikes on hard-shelled prey.

If you would like to hear more about the fast strike of O. scyllarus, including high speed footage of the cavitation bubble and images of damage to the club, I recommend you check out Shelia Patek's TED talk.

References:

Patek, S., Korff, W., & Caldwell, R. (2004). Biomechanics: Deadly strike mechanism of a mantis shrimp Nature, 428 (6985), 819-820 DOI: 10.1038/428819a

Weaver, J., Milliron, G., Miserez, A., Evans-Lutterodt, K., Herrera, S., Gallana, I., Mershon, W., Swanson, B., Zavattieri, P., DiMasi, E., & Kisailus, D. (2012). The Stomatopod Dactyl Club: A Formidable Damage-Tolerant Biological Hammer Science, 336 (6086), 1275-1280 DOI: 10.1126/science.1218764

Breathing for two

Mammals aren't the only ones that give birth to live young. There are many examples of fish, sharks and rays that do it too (just to name the vertebrates!). Some of these also form placentas, like some poecilid fishes and carcharhinid sharks. For those that don't, there's a variety of ways that mothers supply their developing offspring with oxygen, some of them better known than others. A new paper on manta rays provide some new and interesting information about gas exchange between mother and offspring.

Manta rays and their relatives do not form a placenta. Instead, the uterine wall allows gaseous exchange between the mother's blood stream and the amniotic fluid. What wasn't clear, until the new paper, was how oxygen was getting from the amniotic fluid into the developing embryo. The authors from the Okinawa Churaumi Aquarium, in Japan, were fortunate enough to get their hands on a live, pregnant manta ray caught accidentally by fishermen. And, as you do, they placed her in a shallow tank and took an ultrasound of the embryo.

The ultrasound revealed something really interesting. It suggested that the embryo was breathing in a manner quite unlike the way that adult manta rays do. Adult manta rays breath by ram ventilation, that is they push water over their gills as they swim forward with their mouth open. In contrast, the embryo was breathing by dropping the floor of its mouth to suck water in and then closing its mouth to push the water back over its gills, which is known as buccal pumping.

The top panel is a sequence of ultrasound images showing the opening and closing of the mouth in the embryo. The lower panel shows a 3 dimensional images reconstructed from CT-scan data and displayed in the same orientation as the ultrasound images (image taken from the paper).

Sharks and rays that breath using buccal pumping typically have a larger spiracle, a hole near the top of the head behind the eyes. The spiracle acts like a second mouth to help draw water in. It's generally larger in species that live on the bottom, but is reduced or absent in species that are active in mid-water and those that rely more on ram ventilation. Adult manta rays have a spiracle that is reduced to a small slit, but new born manta rays, and presumably embryos, have one that's more reminiscent of an animal that uses buccal pumping to breath.

The spiracle opening of a new born manta ray, Manta alfredi, above and the same structure in an adult below (image taken from the paper)

The authors argue that because the manta ray embryo has no direct connection to the mother that the buccal pumping behaviour must, therefore, be for respiration. And that the rapid loss of the spiracle after birth indicates a shift from respiration by buccal pumping to respiration by ram ventilation. Buccal pumping respiration is also known in embryos of egg-laying species and it may be the dominant type of respiration among sharks and rays that do not form a placenta.

Tomita, T., Toda, M., Ueda, K., Uchida, S., & Nakaya, K. (2012). Live-bearing manta ray: how the embryo acquires oxygen without placenta and umbilical cord Biology Letters DOI: 10.1098/rsbl.2012.0288

Carnival of Evolution #48 - on Pharyngula

Go read up on the best evolution blog posts of the last month at Pharyngula. I really should get around to submitting a post one of these days.

Women in science

Doing science has become very competitive in the last decade or so. Partly because there are so many more people completing a PhD, and partly because of the way research funding is awarded. This is causing many problems for scientists and the process of doing science. For instance, competition for funding has been linked to an increase in the number of papers that are retracted. A new report now links the competitive environment with fewer PhD candidates wanting a career in academic research, especially among women. Read about it in the Guardian.

All the better to see you with

Giant and colossal squid have bigger eyes than any other living animals. A paper published in Current Biology last month asks why it is that they do when other animals get by with smaller eyes. Intriguingly, they suggest that it might be sperm whale predation that has driven both body and eye size in these massive invertebrates. Unfortunately, I think the authors tackle the problem from and adaptationist perspective and do not give alternative hypotheses due consideration.

A kraken fights with a leviathan in a diorama at the Museum of Natural History (image Wikimedia Commons)

Eyes are metabolically expensive organs to build and to maintain. So it's interesting that they are so prevalent in the deep sea where sunlight is weak or never reaches. But, in the deep, animals and other organisms make their own light. Indeed, it has been estimated that 80 - 90% of creatures in the deep sea are bioluminescent. Clearly then, making and detecting light are important for life below the sun's influence. But, most animals get away with eyes much smaller than those of giant and colossal squid.

Generally speaking, big eyes are more sensitive and provide better spatial resolution. But, the amount of light reaching the retina is dependent on the ratio of focal length (the distance between lens and retina) and pupil diameter (camera mavens will recognise this ratio as the f/stop value). In some cases smaller eyes can be more sensitive because they have a shorter focal length. With longer focal length, larger eyes have greater acuity and can increase sensitivity by increasing pupil size.

The paper examines just how much visual performance in the sea improves with increasing eye size and what visual strategies giant eyes are best suited for. They take into account the focal length and pupil size plus a raft of other factors such as the transmission of light in water and background illumination. They find that giant eyes are best suited to detecting the distant shapes of large moving objects illuminated by small bioluminescent organisms disturbed by the object's passage.

The authors argue that the only objects that are both large enough and important enough for giant and colossal squid to detect at distance are hunting sperm whales. The bioluminesence stimulated by moving whales would allow giant and colossal squid to see them at about 120 meters away. Unfortunately for the squid, this is inside the distance at which the whales would detect them with their sonar. The authors argue, therefore, that the squid must use their ability to detect whales at this distance to prepare for a coordinated escape response.

The authors, I'm sure, would acknowledge that their study is speculative and important questions remain. For instance, the maximum recorded sprint speed of hunting whales is about  9 ms^-1, which would give a giant squid more than 10 seconds of advanced warning of a whale closing at speed. Giant and colossal squid are unlikely to be strong enough swimmers to flee beyond sonar range, leaving them the only option of evasive maneuvers. Surely smaller eyes and shorter detection distances would still leave them ample time to prepare to outmaneuver the whale.

The authors rule out the possibility that giant and colossal squid use their eyes to detect prey because huge eyes offer only marginally better performance than much smaller eyes. However, they only consider individual prey items and looking at the published studies on the diet of giant squid, many of their prey species are schooling fish (e.g. Macruronus novaezelandiae, Micromesistius poutassou and Trachurus trachurus) and squid (e.g. Nototodarus sloanii, Ommastrephes bartramii and Todarodes sagittatus). Although detecting individuals of these species might not favour the evolution of giant eyes, schools of prey could easily reach sperm whale size and would also trigger bioluminescence as they move. 

On moonless nights fishermen at the surface are able to use the pattern of bioluminesence stimulated by schools of fish to distinguish among several species. It's possible that giant and colossal squid could use patterns of bioluminescence to determine whether it's being stimulated by prey or non-prey species or, of course, a hunting whale. Longer prey detection distances would seem to be a highly advantageous trait for fueling the fast growth rate of giant squid (reaching 150 - 250 kg in about 5 years).

In making their argument that the eye size of giant and colossal squid is unusually large, the authors contrast them with several other extant whales and fish. None of these extant species, though, are visual predators that hunt at great depth. The authors do also compare to the extinct ichthyosaurs, which had eyes of similar size and were probably visual predators that hunted at depth. They suggest that this is because ichthyosaurs had a similar need to detect the bioluminesence stimulated by large moving objects, perhaps pliosaurs or other ichthyosaurs.

None of these comparisons are truly fair, even the ecologically similar ichthyosaurs, because they don't take into account allometric scaling effects. When making trait comparisons among lineages you should always examine the trait within lineages. It is entirely plausible that the eyes of giant and colossal squid are large simply because they're scaled up versions of those in species with smaller body size. The authors claim that giant and colossal squid eyes are unusually large even for squid, but the paper they cite in support of this point only examines changes in eye size in a single species as it grows.

The problem with looking at the relationship between eye size and body size within a single species of cephalopod is that eye size as a proportion of body size decreases as they age. This is almost certainly not e case when you examine the relationship between body size and eye size among species. As in cephalopods, eye size in vertebrates generally decreases as a proportion of body size as individuals grow. But, eye size increases as you move from species with smaller body size to species with larger body size. Unfortunately, no such among species relationship has been published for cephalopods.

In a crude attempt to get an idea of the scaling relationship in squid, I searched the literature for reports of eye size that could be linked to a length measurement. I was able to find data in three species (Dosidicus gigas, Loligo opalescens and Illex illecebrosus), but eye size as a proportion of mantel length in giant and colossal squid was within the range of these species. Although my "analysis" should be taken with some salt (e.g. some values were estimated from graphs), it seems that giant and colossal squid eyes are not unusually large when their body size is taken into account.

Although I have strong doubts that the eyes of giant and colossal squid were selected for detecting hunting sperm whales, this study does provide some interesting information about the performance of eyes in the sea. Because deep-sea squid and their predators are so hard to observe, we really only have recourse to mathematical models to determine the selective pressures on predators and prey. But, we should never ever start with the assumption that the trait of interest is adaptive and then look for explanations.

Nilsson, D., Warrant, E., Johnsen, S., Hanlon, R., & Shashar, N. (2012). A Unique Advantage for Giant Eyes in Giant Squid Current Biology, 22 (8), 683-688 

DOI: 10.1016/j.cub.2012.02.031