Continuing the series, let us now take a look at one weird turtle species in particular: Dermochelys coriacea, the leatherback sea turtle.
While the utter weirdness of D.coriacea is ultimately the main reason for why it wound up in this series, there is an ulterior motive. Having searched the internet for general information on the species I found myself rather disappointed with the amount of utterly generic / wrong info regarding leatherbacks. Its Wikipedia entry is particularly disappointing. So here’s hoping this influx of information can help alleviate that.
A turtle without a shell?
Yes, it’s true, leatherback turtles have lost their shells. Shell reduction is relatively common in turtles. It seems a little funny. After going through all the trouble of evolving impregnable armour, many taxa then went out and removed large chunks of it. We see shell reduction in snapping turtles (Chelydra and Macrochelys), soft-shelled turtles, and even other sea turtles. None of them, however, reduced their shells to the point of actually removing them.
In leatherbacks the “shell” is nothing more than a loose collection of osteoderms spread over the back and belly. There is no longer a definitive carapace, or plastron. In fact leatherbacks don’t even produce Beta-keratin (the hard component of reptile scales). Instead this has all been replaced by thick, leathery skin.
Biggest of the big starts off tiny and grows fast
D.coriacea has the distinction of being the largest living turtle alive. Individuals average around 680 kg (1,496 lbs). The largest individual ever recorded reached a whopping 916 kg (2,015 lbs) in mass (Eckert & Luginbuhl 1988). “Shell” length is approximately around 2.3 m (7.5 ft), with a front flipper length of up to 2.74m (9 ft). This puts leatherback turtles in the same league as many pinnipeds (seals & sea lions). However, unlike pinnipeds which start off life as 23-35kg animals (1/37th and 1/17th average adult body size respectively), hatchling leatherbacks start off as 40-50 gram animals (Davenport et al, 2009)!
That means over the course of leatherback ontogeny, these animals grow over 4 orders of magnitude, or 10,000 times their birthweight!
Even more surprising than that is just how fast leatherbacks can reach sexual maturity. Traditionally sea turtles were thought to fall in line with the Aesop fable and grow “slow and steady” throughout life. This has lead some researchers to estimate ages up to 50 years for some sea turtle species to reach sexual maturity.
Leatherbacks don’t seem to fit this mold at all. Turtle researcher Anders Rhodin looked at the chondro-osseous morphology of leatherback limb bones, and from the apparent rate of cartilage absorption estimated that individuals may reach adult size in as little as 2-3 years (Rhodin 1985)! To put that in perspective it would be like a human baby growing from 3.2 kgs (7 lbs) at birth to a 68 kg (150 lbs) adult in just under 5 months!
Skeletochronology studies done since 85 have cast doubt on Rhodin’s initial 3 year prediction; pushing adult size out from 3 years to 5, 9 and even as much as 30 years (Zug & Parham 1996, Jones 2009, Avens et al 2009).
However, these authors all used caution with their findings and pointed out several possible flaws that could affect the study; the most important of which was the assumption that lines of arrested growth (LAGs) in turtle bones, reliably indicate an annual cycle. Bjorndal et al (1998) called into question the use of skeletochronology for assessing sea turtle age when their comparison of annual bone growth (using a tetracycline marker) found a complete lack of LAGs in their study species (green turtle Chelonia mydas). The authors called into question the assumption that LAGs are an endogenously produced phenomenon (independent of environment) and could reflect different cycles that may occur multiple times a year, or more infrequently (e.g. El Niño/La Niña events).
Recent reviews of leatherback growth data (Snover & Rhodin 2008), have found that even if leatherbacks do not reach sexual maturity in 3 years, they do seem to exhibit growth rates on par (and sometimes faster) that comparably sized mammals. Further, there has been rumoured talk of new research that was shown in a herpetological conference last year that suggests that Rhodin’s initial estimate of 3 years might be right after all.
In most tetrapods, a thin layer of cartilage exists on the epiphyses (ends) of the limb bones. New cartilage forms from the top while old cartilage (closer to the metaphysis, or body of the limb bone) gets replaced with bone. Collectively this endochondral ossification leads to bone and body growth. In mammals, birds and most squamates, as the animals reach sexual maturity, the rate of cartilage growth gets outpaced by the rate of bone replacement until these epiphyseal plates close.
This works the same in turtles, except that the epiphyseal plates never fully close, thus allowing the potential for continuous growth throughout adulthood (though growth rate much past sexual maturity is often so slow as to be negligible).
An interesting aspect of cartilage is that it tends to grow slowly. This has nothing to do with cellular metabolism (mammal cartilage and shark cartilage grow equally as slow), but rather with the way in which chondrocytes (cartilage cells) get their nutrients. All nutrients come to cells via the blood, but blood flow to chondrocytes is often dispersed and indirect. Most chondrocytes rely on the diffusion of nutrients from the blood stream through their watery cartilage matrix. This diffusion puts extreme limits on the rates at which cartilage can grow.
Consider how long it takes for a cut on your ear to heal, compared to a cut on your arm (and by cut on the ear, I mean anywhere that isn’t your earlobe — which is composed of fat and not cartilage).
There are ways that animals have found to get around the slow growing cartilage problem. Mammals, birds and varanid lizards have added extra blood vessels that surround the cartilaginous epiphyses, allowing for nutrients to more readily diffuse towards the chondrocytes. Leatherbacks, however, have taken things a step further. Along with blood vessels that surround the chondroepiphyses, leatherbacks have added a second set of blood vessels that actually cut through the epiphyses, allowing for a much closer association of chondrocyte to nutrients (Rhodin et al 1981). This type of transphyseal circulation has also been found in whales and pinnipeds; both groups which are known for phenomenally fast growth rates.
Leatherbacks grow big by eating crap
So leatherbacks start life off small, and they grow really big in a short amount of time. One would suspect, then ,that they must be eating a high energy, high protein prey item. Well…not so much. D.coriacea makes its living by solely eating jellyfish. At 95% water, jellyfish aren’t exactly a high energy food source; especially when compared to more typical fare like krill — which are used to sustain baleen whales. The energy content in jellyfish are so low that it has been estimated that growing leatherbacks would need to eat their (ever increasing) bodyweight in jellyfish, daily (Lutcavage & Lutz 1986), while adults consume ~50% of their bodymass per day just to cover maintenance costs (Davenport & Balazs 1991). Not only are jellyfish 95% water, but they are isotonic too, so they are 95% saltwater. No surprise then that leatherbacks also possess the largest salt excreting glands of any sea turtle. Leatherback salt excreting glands are 1.6 times the size of their eyes, and 1.7 times the size of their brains (Hudson & Lutz 1986).
Leatherbacks are marathon swimmers
Eating 200kg of jellyfish a day is no mean feat. While jellyfish do tend to travel in schools, they seem to rarely be found in the kind of huge “blooms” that one can see with krill. This means that a foraging leatherback has to spend a lot of time on the move. No surprise then that data on wild D.coriacea show that they spend between 93% and 99.9% of any 24 hr period actively swimming around (Eckert 2002). These turtles don’t slowly cruise around either. Foraging leatherbacks can cover 54 km per day, showing a bimodal distribution of swim types. Eckert referred to them as “type 1” and “type 2” swim behaviours.
Type 1: This was a high speed swim that slowly decreased over a period of 5-15 min, and then increased again.
Type 2: This was a continuous high speed swim.
Turtles averaged around 2-3 km/hr during type 1 swim behaviours, while type 2 behaviours saw speeds up to 10km/hr.
Both speeds are all the more impressive given that this is in the far more dense medium of saltwater, rather than air.
Leatherbacks have a quantized distribution
Easily one of the most well known facts about D.coriacea is that it is the poster child for gigantothermy. This is the concept proposed by James Spotila and others (Spotila et al 1991) as an alternative to automatic endothermy for dinosaurs. I have explained gigantothermy before on other portions of the site, but in brief: gigantothermy is the phenomenon in which large bradymetabolic animals can retain their metabolically generated and externally acquired heat by the shear virtue of having a much larger volume compared to their overall surface area. As heat from the body leaves via conduction, having less skin in contact with the air/ground/water, means a smaller “window” for heat to escape. Thus a gigantothermic animal can have all the benefits of being warm all the time, without the side effects seen in automatic endotherms (i.e. devoting >90% of all food resources to staying warm). Gigantothermy will be touched on more below. For right now I want to concentrate on just one particular aspect of it.
So the benefits of gigantothermy is the greater thermal lag present in larger body sizes, but as mentioned earlier, leatherbacks start off life at substantially smaller body sizes. Much like with the arguments against gigantothermic dinosaurs, a baby leatherback would have none of the advantages of gigantothermy, and so would be expected to show a much narrower thermal range than an adult.
Researcher Dr. Scott Eckert set out to test this assumption by combing through the literature in search of all documented accounts of juvenile leatherback sea turtle sightings. An aggregation of 100 different published sources showed an interesting development with leatherbacks. While hatchlings did indeed show a much lower thermal tolerance (hatchlings < 100 cm in curved carapace length were always found in waters 26°C [79°F] or warmer), once turtle hatchlings reached a certain — rather small — size of 1 meter, they could be found in waters as cold as 12°C (53°F). In other words, leatherbacks don’t show a gradual increase in thermal tolerance as one would predict by the gigantothermy model, but rather a quantized distribution, with turtles going from a narrow thermal tolerance to a huge thermal tolerance, once they hit a specific size. Exactly what that size is, is not known, but judging from Eckert’s findings, it is probably very close to 1 meter in curved carapace length (Eckert found data on juveniles with curved carapace lengths of only 1.07m, inhabiting 12°C water).
How is this even possible? Well there are probably two unique things in play here.
Leatherback muscle is temperature independent!
One unique feature discovered about the propulsive muscles of leatherbacks is that, unlike most other vertebrates, their muscles fire at the same rate, and strength, regardless of the external temperature (Penick et al 1998).
In a preliminary experiment Penick et al lowered the body temperatures of two adult D.coriacea and measured their metabolic rate during this time. The authors discovered no appreciable drop in metabolism with such a rate drop. This thermal independence of metabolism got the authors interest, so they biopsied pectoralis muscle from nine nesting leatherback females. The excised muscle was then subjected to seven different temperatures spanning 35° (5°-40°C). The authors measured the oxygen consumption of the muscle during all trials.
The results were astonishing. From 5°-35°C the pectoralis muscle of D.coriacea showed complete thermal independence. As controls, the authors used biopsied muscle tissue from green turtles (C.mydas), spiny tailed iguanas (Ctenosaurus similis) and marine toads (Bufo marinus). All the control animals showed a thermal dependence of muscle metabolism, but the leatherback pectoralis muscle exhibited complete thermal independence over almost the entire range of temperatures that wild turtles would experience. That’s not just weird for a reptile, that is weird for an animal!
In most animals, muscle performance is intimately tied to temperature. Consider how hard it is to hold a tight grip when your hands are cold. Colder muscles fire at a slower rate, and often with less motor units (i.e. cold muscles tend to be weaker too). Logically this makes sense. Muscle contraction is ultimate a multi-step biochemical interaction, and heat (temperature) is a measure of how much energy is present in a particular molecular system. The more heat, the more energetic the reaction can be (up to the point of denaturing). Most cellular processes work best within a certain temperature range. That range is often pretty low (1-2 degrees). Once out of the molecular “sweet spot,” reactions become less efficient. Automatic endotherms (mammals & birds) deal with this problem by making sure most of their cells stay within this “sweet spot” all the time. Their muscles are still temperature dependent, but they are housed in a protective “bubble” of heat that allows them to operate under a range of conditions.
Many ectotherms — too — have found a way to circumvent the molecular “sweet spot” problem. If one were to measure the metabolic rate of a thermoconformer like a snail living in the intertidal zone (where behavioural thermoregulation is nigh impossible), one will find that the range of temperatures at which the metabolic rate of the snail stays the same, is much larger than one would expect given the usually 1-2° limit of many biochemical reactions. The reason for this apparent thermal independence is due to a change in the chemistry of the chemical reactions. For instance, enzymes that would normally catalyze reactions at one temperature may be altered to show a higher affinity for their substrate at lower temperatures. In so doing, the hypothetical thermoconforming snail is able to keep a constant metabolic rate in the face of rising, or lowering temperatures; at least within limits.
The pectoralis muscle of leatherbacks takes this enzyme affinity idea a step further, and covers a much larger range of temperatures than any other animal so far measured. As the authors wrote:
Temperature independence of leatherback pectoralis muscle metabolism is a rare example of perfect metabolic
compensation of tissue metabolic rate over the range of environmental temperatures experienced by a vertebrate.
The metabolic stability of resting leatherback muscle may facilitate survival in both cold and warm water, allowing use of a broad thermal niche, not only by adaptations for heat retention, but also by direct regulation of tissue metabolism, such that pectoralis muscle functions independently of its thermal environment.
So that may be one way in which juvenile leatherbacks are able to quickly inhabit a broad thermal range. However this is not the only way in which D.coriacea is able to pull this off. Leatherbacks have another interesting thermal secret.
Leatherbacks have brown fat!
Brown fat is a special kind of adipose (fat) tissue that is seen in many birds and mammals. Unlike white/yellow fat, brown fat is not a form of energy storage; at least not usable energy. When brown fat is burned, 100% of the energy released is in the form of heat. This 0% efficiency in usable energy is actually beneficial to mammals because it can allow an animal to increase its body temperature without having to resort to shivering. Brown fat distribution in the body varies from species to species, and is often only seen in the young of mammals (e.g. human babies have extensive brown fat deposits around their necks). Brown fat has long been associated with automatic endothermy, so the discovery of brown fat in a reptile was rather unexpected.
This surprising bit of little known information is actually remarkably old. It originally started with a prediction by herpetologist Nicholas Mrosovsky (1980). Mrosovsky predicted that the ability of D.coriacea to maintain high body temperatures in cold waters might be more than a factor of large size and extensive fat deposits. He hypothesized that some of those fat deposits might actually be brown fat. Eight years later researchers Gregory Goff and Garry Stenson dissected leatherback specimens that died in fishing accidents. They discovered that D.coriacea has two layers of fat. The most superficial layer consists of the classic white, solid fat that is seen in many animals. Deep to this layer, though, was a distinct second layer that consisted of softer tissue that was brown in appearance. Besides the colour this tissue was easily distinguishable by the extensive vascularization present throughout. The two layers were separated from each other by extensive sheets of fibrous connective tissue. This brown layer was found to encompass the back of the neck, the inguinal region, parts of the plastron, and to lay over the vertebrae of the carapace. Leatherbacks don’t just have brown fat, they have a tonne of it.
Though this was the first time that anyone had ever found brown fat in a reptile, it was not the first time that brown fat was discovered in a bradymetabolic animal. Prior to this brown fat deposits were found in xiphiid fish (swordfish and kin) around their eyes. This brown fat was believed to be responsible for allowing swordfish to keep their eye temperatures 10°-15° C above ambient water temperatures. It is likely that the extensive presence of brown fat in leatherbacks plays a significant role in their thermogenic abilities and may allow young, small individuals to colonize habitats that would normally be too cold for them.
But wait, there’s more.
Inspired air is rarely ever close to the temperature, or humidity of the body. Creatures that have a high rate of respiration, or live in very arid environments, run the risk of getting dehydrated as body moisture gets pulled out by drier air. It has been hypothesized that respiratory turbinates — which are long thin scrolls of bone, or cartilage that contain well vascularized tissue — evolved in response to this air conditioning problem (Hillenius 1992). By keeping body warmed blood close to the inspired airway (the nose), cooler air has a chance to gain heat from the blood, and approach core body temperature before reaching the lungs. More importantly, the expired air that is now at body temperature has a chance to lose that heat to the progressively cooler blood found towards the extremities of the nose, and in the process of losing its heat, cause much of its acquired moisture to condense on the turbinates, thus saving body water from being lost to the environment.
The extensive presence of respiratory turbinates in mammals and birds has been used to suggest that these anatomical peculiarities could be used as a means of detecting the thermophysiological status of extinct taxa; namely therapsids and dinosaurs (Bennet & Ruben 1986, Ruben et al 1996, Hillenius & Ruben 2004). This hypothesis has been looked at controversially by many in the paleontological community (though surprisingly few rebuttals actually appear in print). One of the oft cited counters to this hypothesis is the lack of universal presence of respiratory turbinates in both mammals and birds. Many of the case studies presented are from aquatic mammals, or birds, making them of questionable value. If one is living in an aquatic environment and are breathing infrequently, then one would expect respiratory turbinates to no longer be selected for. Then there is the “no reptile has respiratory turbinates” argument, that is currently lacking the cross taxa exhaustive study that would really be needed to back up such a statement.
But I’m digressing.
Last year, a group of researchers in Ireland found two dead leatherback sea turtles that had washed ashore in Ballycotton, East Cork Ireland. The animals were large females (450kg, or 990lbs). Both animals were dissected, and CT scanned (Davenport et al 2009). The researchers discovered an interesting association of blood vessels around the trachea of adult leatherbacks. When they compared it to a gross dissection of a hatchling, they noticed that the blood vessels surrounding the trachea did not exist in the youngster. Further, unlike other sea turtles, or other tetrapods for that matter, the trachea is not covered with a interspersed collection of cartilaginous rings, but is rather encased in a sleeve of uncalcified cartilage. The authors speculate that having a completely cartilaginous covered trachea would allow leatherbacks to dive to great depths (up to 1280m [4198ft]. See Doyle et al 2008) that would progressively crush the airway, but allow for elastic rebound of the trachea upon reaching the surface. That this tracheal architecture only appears later in ontogeny is very suggestive of its functional role in diving.
Then there was the somewhat bigger news. Davenport et al discovered a well developed plexus of blood vessels along the trachea. These blood vessels were longitudinally arranged and found in the deeper 2/3rds of the mucosa, with numerous cross sections connecting them. The presence of this extensive vascular network along the trachea suggests the presence of a counter-current heat exchanger in the airway. In other words, Davenport et al appear to have found something equivalent to a respiratory turbinate, but in a reptile.
A logical question one might ask at this juncture is: “does a turbinal structure really make any sense in a marine animal?”
Studies on leatherback breathing show that the entire respiration process takes all of 2 seconds to occur, and that breathing intervals of 1 breath every 13 seconds are common when swimming near the surface. Given that respiratory turbinates are supposed to function in maintaining body temperature and humidity, will a counter-current heat exchange system have any real effect during this kind of respiration process?
The answer at the moment appears to be unknown. While it is doubtful that this heat exchange system is doing much to maintain body moisture (given the environment leatherbacks inhabit, it hardly seems a problem), it may still serve a vital air warming role. The lungs of a leatherback swimming in the waters around Nova Scotia will be at a core temperature that can be as high as 18° C above ambient water temperature (Frair et al 1972 measured turtles in 5° C water, but adults are known to dive in 0° C water as well [James et al 2006], so the temperature differential is likely even greater). The difference between the core temperature and the ambient air would be even greater (as high as 32° C). Such a large temperature differential between the body core and the air, coupled with the rapidity at which the air is inspired could cause a shock to the delicate lung tissues. By having an extensive vascular network in the trachea, the rapidly inspired air (which can be as low as -9°C in January) has a chance to reach core body temperature before reaching the delicate lungs, and possibly causing a chelonian version of pneumonia.
An alternative explanation that Davenport et al suggest is that this extensive vasculature may instead function to help “inflate” the trachea after long dives. This suggestion is based off of the finding of Cozzi et al (2005) of a similarly well vascularized structure in the striped dolphin (Stenella coeruleoalba). Interestingly enough, cetaceans are one of the exceptions to the “all mammals have respiratory turbinates” rule.
Davenport et al express their doubts over the “inflation hypothesis” stating that:
We believe that the elasticity of tracheal cartilage (combined with expansion of the remaining tracheal air during ascents) will be sufficient to reinflate the trachea, and that the vascular lining of the dolphin trachea also helps maintenance of respiratory temperature and water balance.
One thing is for certain, given the extensive vasculature seen in the trachea there is almost no way that it could not function as a counter-current heat exchanger. Whether, or not that is its primary function remains to be elucidated.
The “warm-blooded” turtle that wasn’t
Given their remarkable ability to maintain high body temperatures, coupled with their very aerobic lifestyles, high growth rates, presence of thermogenic brown fat and even the possibility of a tracheal version of respiratory turbinates, it really seems like leatherbacks are reptiles that have converged on automatic endothermy just like mammals and birds.
No surprise then that many scientists have expressed interest in the metabolic status of these incredible animals. Probably the most oft cited study in this arena was that of Frank Paladino and others (1990). These authors measured the metabolic rates of female leatherbacks while covering their nests, or crawling to/from the water. The results of their study found, surprisingly that leatherback resting metabolic rate was three times higher than expected for a reptile of that size. This suggested to the authors that, while leatherbacks were not automatic endotherms, they were on their way.
Studies since then have relooked at the data from Paladino et al and found conflicting results (e.g. Lutcavage et al 1990, Paladino et al 1996, Davenport 1998, Wallace et al 2005). A fantastic rundown of the current state of leatherback thermophysiology can be found in Wallace & Jones 2008. The authors give a nice rundown of the myriad ways in which metabolic rate can be measured / inferred, along with the advantages and limitations of each technique.
In order to get to the bottom of these conflicting data, Wallace & Jones compared allometric relationships between body size and resting metabolic rate for green turtles (C.mydas), leatherbacks (D.coriacea) “typical reptiles” and mammals. Their results were telling. Leatherbacks and green turtles did show a higher mass specific metabolic rate than “typical reptiles,” but the reason behind this has more to do with methodology than actual physiology.
Few allometric studies take into account changes in metabolism from hatchling through adulthood in most reptiles; especially large bodied reptiles that change body size by multiple orders of magnitude throughout ontogeny. The allometric study the authors used (Bennett 1982), suffered from a lack of data for reptiles that were anywhere near the body size of adult sea turtles. To help alleviate this problem the authors only compared the slope and intercept values from their allometric equations for green turtles and leatherbacks, to those of the generalized reptile and mammal study.
A statistical assessment of the data revealed that the allometric relationship between body mass and resting metabolic rate for leatherbacks and green turtles was not significantly different from that of the generalized reptile equation, but was significantly different from that of the mammalian equation. If anything the data suggests that the constant used in Bennett’s (1982) equation was too small (likely due to the choice of animals used).
Which brings us to the punchline; the weirdest thing about about leatherbacks is that despite their speedy growth, highly active nature, thermal independence, heat generating organs, and respiratory turbinate-like tracheal structures, leatherbacks are still bradymetabolic animals. Wallace & Jones cited this as a possible phylogenetic constraint, but I would beg to differ. Dermochelys coriacea has found a way to have all the benefits of being warm all the time without having to devote most of its available energy to a wastefully high metabolism.
~ Jura – who was able to avoid using a single picture of a sea turtle hauled out on the beach.