"White-Blooded" Icefish, 1927

A bizarre group of Antarctic fishes lost their red blood cells but survived to tell their evolutionary tale, revealing a fundamental lesson about the birth and death of genes.

By Dan Cossins | April 1, 2013

CURIOUS CATCH: Norwegian zoologist Ditlef Rustad (left) hauls his net out of the frigid waters of the Southern Ocean. Part of an expedition to Bouvet Island in 1927, Rustad discovered a bizarre fish with a protruding jaw, creamy white gills, and transparent blood. He called it the “white-blooded crocodile fish.” PRIVATE COLLECTION, LIV S. SCHJERVEN

On December 1, 1927, zoologist Ditlef Rustad pulled ashore on Bouvet Island, 1,750 kilometers off the coast of Antarctica, as part of a Norwegian expedition to claim the remote, wind-whipped island as a whaling outpost. Later that month, casting nets into the frigid waters, Rustad hauled up a very strange-looking fish. It had no scales and was very pale, even translucent in parts. Behind its protruding, crocodile-like jaw, he saw gills that were milky instead of the usual crimson. And when Rustad cut open the fish, he saw that its blood was transparent, like ice water. “Blod farvelöst,” he wrote in his notebook—“colorless blood.”

In a 1954 Nature paper, biochemist Johan Ruud confirmed that Chaenocephalus aceratus lacked red blood cells and hemoglobin, the protein that carries oxygen around the body and gives blood its red color. “It was a shocking discovery,” says William Detrich of Northeastern University, who has spent most of his career studying C. aceratus and the other 15 recognized species in the family Channichthyidae, or the Antarctic icefishes. “Among the 50,000 or so species of known vertebrates, these fish are the only examples that lack both hemoglobin and red blood cells.”

COLORLESS BLOOD: Now known as Chaenocephalus aceratus, or the blackfin icefish, Rustad’s catch is one of 16 recognized species of Antarctic icefish—the only vertebrates known to lack hemoglobin and red blood cells. Thanks to the extremely high oxygen content of the Southern Ocean, the icefish appear to have survived the loss of globin genes long enough to make compensatory adaptations.PRIVATE COLLECTION, LIV S. SCHJERVENInstead, icefish rely on oxygen dissolved in their plasma, which can transport only 10 percent as much oxygen as the hemoglobin-filled blood of most fish. Presumably to compensate, they developed much larger hearts and blood vessels, far denser capillary beds, and a much higher volume of blood. But “the icefish went through a period where they weren’t particularly well adapted,” Detrich notes. Indeed, they likely only survived the initial loss of hemoglobin as a result of the extremely high oxygen levels in the waters of the Southern Ocean—and the fact that they had so few competitors or predators.

Comparing icefish DNA with that of red-blooded fishes in 2006, Detrich identified the remnants of the two globin genes required to make hemoglobin. Since these genes were not entirely necessary for the survival of the icefish in their oxygen-rich sanctuary, Detrich hypothesizes, mutations could accumulate that rendered the genes nonfunctional without killing the fish. “Evolution is a tinkerer,” Detrich says. “It works with what’s available, and if what’s available isn’t needed, it gets thrown out.”

Though it is not the only example of this fundamental rule, the bizarre white-blooded fish hauled up by Rustad in 1927 is surely among the most extreme cases—a stark illustration of how even genes thought to be absolutely essential for vertebrate life can, in the right conditions, be allowed to perish. 

Watch a short film from the Howard Hughes Medical Institute about how scientists have pieced together the evolutionary history of the Antarctic icefish.

Add a Comment

Avatar of: You



Sign In with your LabX Media Group Passport to leave a comment

Not a member? Register Now!

LabX Media Group Passport Logo


Avatar of: Roy Niles

Roy Niles

Posts: 116

April 8, 2013

Evolution is a tinkerer, yes, but then it takes intelligence to decisively tinker.

Avatar of: jeenious


Posts: 45

April 11, 2013

As a science literate thinker, and as a thinker in search of logical consistency, I find many implied contradictions, even in peer review papers, concerning what "evolution" is, and what "evolution" does (or, whether it can be soundly argued that it "does" anything at all).

Often, when I communicate my self-honest, scientifically valid, questions about things stated and/or implied about evolution of species, I am mistaken for someone having an ideological or doctrinal agenda in mind. I have nothing of the kind. I merely see direct contradictions -- some stated and some implied -- in writings about evolution from inside the life sciences, as well as from without.

Here are a few representative examples of questions I have:

Does evolution,"bring about" a particular change? Or would it not be more scientifically consistent to say that evolutionary changes are merely serendipitous "results?"

Compare to the question in climate science, whether a tornado "brings itself about" or whether it is a "result" of the coming together of certain set of contributory happenstances.

Does evolution "read" what is going in with a species's physiology in a given place over a given period of time, "detect" an opportunity (such as a food chain niche that is not being availed by itself or by other species) or a survival-challenge (such as an on-going decline in average temperature), "discern" which physiologic sub-system or set of sub-systems would be the one or ones most likely to be able to come up with a mutation (or a complex package of mutations) likely to provide a solution, "plan" one or more mutational experimentations, "implement" that trial solution, or those ones, "detect" accomplishment of -- or progress toward -- which mutations succeed and which ones do not succeed. OR, isn't it more likely that mutations are merely random, and serendipity is the one and only game in town?

FREQUENTLY wording in articles about evolution (or which allude to evolution peripherally in discussion of another subject in a life-science) attributes the development of a particular change in a species (or into the origin of a species) to the "need" for such a change. Wouldn't it be more scientifically consistent -- if we would not wish to attribute "intelligence" to "evolution," if we would use words that say, and imply, that a species serendipitously mutated in a fortuitous way? Or, in a highly complex and ramified package of ways that are, by hindsight only, effective in providing survival or in allowing a particular species to prevail reproductively.

These are legitimate scientific questions. They are not doctrine-driven, ideology-driven, agenda driven... EXCEPT that they are driven -- as all scientifically valid questions are -- toward resolving conflicts among our interpretations of empirical evidence.

Sometimes, when I ask such questions, there will be those who respond by arguing that the use of such metaphorical literary devices as personification, analogy, simile.. should be no less helpful to scientists, or to clever writers about science, than to any other class of communicators.

My response to that is this: Literary devices are helpful only so long as they are the TAIL wagged by the DOG of sound reason. Much confusion reigns in discussions of the subject of evolution, or evolutionary theory, as a result of the fact that the DOG of descriptive and analytical accuracy often is wagged to the point of total distortion and scientific incoherence by the TAIL of carelessly and irresponsibly and sloppily applied metaphor.

Avatar of: jeenious


Posts: 45

April 11, 2013

In the article appear these words:


“Evolution is a tinkerer,” Detrich says. “It works with what’s available, and if what’s available isn’t needed, it gets thrown out.”

Though it is not the only example of this fundamental rule, the bizarre white-blooded fish...(etc.)


It cannot be had both ways. Either evolution has qualities of being able to "do" something, "read" something, "discern" something, "respond" to that discernment, make that response "appropriate to a given opportunity or need," "decide" to mutate at one or more RNA loci, "follow through" with a solutional process, "purposefully delete" non-advantageous RNA configurations at those loci... and do more steps in tinkering... OR it does nothing, and the optimization of survival or opportunistic availance is purely by chance alone.

Again, no consistent scientific model can have it BOTH WAYS.

Thus we must choose WHICH of two inter-contradictory postulations we wish to hold:

A. That mutations are random, and fortuitous changes in morphology which, when fortuitous are conducive to fitness to survive, or they do not; OR,

B. That mutations are subject to the super-imposition of a capacity of biological reproductive processes to read, interpret, respond appropriately to, edit and improve upon the morphology of a species in a proactive manner.

Take your pick between these two, OR, any scientific interpretation of evidence DEMANDS an accounting for the mechanism for each FUNCTION whereby the reading, the interpreting, the pattern analysis, the planning, the implementation, and the monitoring of results is attained.

And, what makes a "proactive view of evolutionary dynamics" all the more difficult to support scientifically, is the logical requirement that no one of these listed FUNCTIONS can be lacking in hard evidenciary substance, or the whole chain of ASSUMPTIONS fails to work as a "proved" model.

Let me say that it would please me very much to see empirical support for each and every one of these ESSENTIAL functions occurring in evolutionary change. We should never simply ASSUME or PRESUME into place an interpretive model we cannot (as yet) support empirically, and take a DOGMATIC stance that evolution is a proactive process (as opposed to a merely serendipitous set of patterns in results) until proven otherwise.

To say, or imply, that evolution (as a force, or an entity, or a mechanism in and of itself which proactively shapes the flow of natural changes over successive generations of species) is to state or imply false accuracy.

And, most poignantly in the instant case, to assert in a matter of fact kind of way that "evolution" does something proactive, such as "tinker," or "throw away" anything, and then to assert that such is a "rule," is scientifically unsound.

In science, as much if not more than in any other intellectual endeavor, we need to take care that we do not mislead, or misinform, or make claims based on convenient or easy-to-understand -- but inconsistent or unsubstantiated --pseudo-explanations of yet-undetermined "rules."

Avatar of: Roy Niles

Roy Niles

Posts: 116

June 16, 2013

So the argument is that we must assume that evolution occurs by chance until we have found uncontroversial evidence that it doesn't?  Couldn't we then just as well assume that it occurs because it was predetermined in a chanceless universe?

And which of those assumptios would more likely explain the consistent nature of biological change?  Or do we perhaps need a third assumpton, such as that life is a process that takes advantage of accidental change?

July 24, 2013

Wow, this is amazing!

Popular Now

  1. How to Separate the Science From the (Jerk) Scientist
  2. Could a Dose of Sunshine Make You Smarter?
  3. Prevalent Form of Childhood Leukemia May Be Preventable
  4. Arizona Moves to Alter Wording About Evolution in Education