Bloody Mosquito Redux

In mid-October I started a list in a google spreadsheet of “Articles of Interest.” It’s already up to 162 entries right now, and contains news articles and blog posts from a wide variety of sources that I might need to find again. The news stories behind a couple of recent ICR articles – on Dmanisi and the ancient galaxy from Monday – feature, as do many other topics which are likely to come up. The ICR tends to operate on a time delay of a few weeks, so looking forwards in the seven days following the galaxy story we have new information on Titan’s lakes; some abiogensis research; better dating for Homo (erectus?*) rudolfensis; and also that silly “Junk DNA face” story. Looking back at things they seem to have missed, meanwhile, turns up items like a story about blue straggler stars, a topic about which Brian Thomas has previously made noises, and an interesting system of extrasolar planets.

My point here is that there is no shortage of fresh science news of the kind the ICR likes to talk about. And yet for the second time in as many weeks Brian Thomas has decided to revisit an older story in order to better make a fool of himself. For Wednesday we have Questionable Dating of Bloody Mosquito Fossil – my previous post on this can be found here.

Researchers recently examined a spectacular mosquito fossil from the Kishenehn Formation, finding fresh blood—remnants of its last meal—still stored in its abdomen. They wrote, “The data reported herein provide incontrovertible documentation of the presence of heme- and arguably hemoglobin-derived porphyrin in a 46-million-year-old compression fossil.” How did they obtain this age for the fossil, and are their methods and conclusions reliable?

An earlier Creation Science Update news article confirmed the fossil’s genuine blood remnants. It also introduced the biochemical decay rate experiments that give compelling reasons for excluding hemoglobin from materials that could possibly last for even one million years.

The original story was that a mosquito fossil had been found with evidence of a bloody meal inside of it: this proves that blood-eating mosquitoes have existed for at least the last 46 million years. To Brian Thomas, however, this was evidence that the Earth is young because haemoglobin could not last for 46 million years. The problem is of course that they did not find any haemoglobin. Instead, they found a variety of porphyrin molecules that appear to be the decay products of heme, which is often associated with the much larger and potentially more fragile haemoglobin molecule in the blood. But despite the common association heme is not haemoglobin, and so saying that haemoglobin could not have survived “even one million years” is of no relevance to this story.

More generally, the entire soft tissue/biomolecule preservation topic rests upon a single issue: what do we trust to tell us how old a fossil is? Fossils are usually dated via a variety radiometric methods, which creationists despise because they provide numbers far in excess of 6000 years. These techniques are considered by scientists to be pretty reliable, being insured by the underlying physics of the universe: it may be possible to speed up the decay of unstable isotopes, but it would require a god out to intentionally trick us. In contrast young Earth creationists like Thomas often point to the presence of proteins in soft tissues in a few select fossils as evidence that they must be only thousands of years old. But unlike radiometric dating, the lab studies that purport to show how long these molecules can last are notoriously unreliable. For example last year we saw a study attempting to test how DNA decays in the real world: one finding that Thomas didn’t mention is that they found that DNA can last hundreds of times longer in the field than laboratory testing had predicted. For other molecules these lab tests are all we have – who knows just how wrong they are.

Radiometric dating is the only possible winner in a contest between these two “dating” methods. If the isotopes show that the fossil is old, but it contains soft tissues that may imply youth, then it is the isotopes you should trust. This is not a situation that young Earth creationists like, and so today Mr Thomas is going to try to attack the radiometric dating of our mosquito fossil in order to try to level the playing field somewhat. He begins:

Publishing in Proceedings of the National Academy of Sciences (PNAS), the authors of the mosquito fossil report did not independently test the fossil’s age but cited a previously published age given to the Kishenehn Formation.

The cited publication listed two “age” results for the relevant layers within the Kishenehn: 46.2+0.4 and 43.5+4.9 million years. Why the PNAS authors chose 46 million years, an age based on radioisotope dating involving argon decay, over other dates ranging within the error span (38.6 to 46.6 million years) was not explained.

When Thomas has a + here he really means ±, signifying that there is a margin of error “plus or minus” 0.4 million years (or 4.9 in the second case). In short, the fission-track dating technique had dated a specific layer of tephra to between 38.6 and 48.4 million years old, while the argon-argon method gave a more precise age for the same layer at 45.8 to 46.6 million years. Here is a number line, for those of you that prefer to visualise your numbers: red is fission-track, blue is argon-argon.

As you can see, the range of dates produced via the fission-track technique completely encompasses the argon-argon results from the same tephra layer – something which would be quite unlikely if the processes that YECs claim render these methods unreliable had ever been in operation. Because they are dating the same thing the place where the two bars overlap – the entire range of the argon-argon method – is the most likely value. Now, the authors of the mosquito paper would not have been particularly concerned about dating their fossil to the precision of hundreds of thousands of years (it being tens of millions of years old was really enough for them), so an age to two significant figures would be more than enough. What value would you round to other than 46 million years? 47 is possible, on the grounds that the upper bound extends to 46.6, but would be a much more arbitrary choice.

Brian Thomas, a former schoolteacher, appears to have no conception of how to properly deal with error bars and rounding. Being in his class sounds like it would give you good grades – you could give him any value in an experiment you liked, as long as it was within the margin of error of at least one measurement attempt – but would be unlikely to actually teach you much.

He’s got more for us today:

Argon-argon dating begins by measuring relative amounts of argon isotopes in an igneous rock sample. Though the Kishenehn is largely sedimentary, not igneous, it contains some igneous material that holds argon. Researchers generated age estimates, assuming 1) that no argon had entered or exited the material after a volcano deposited it and 2) that all of the argon was only one of the two possible isotopes when it was initially deposited.

However, independent studies have debunked the second assumption. For example, one geologist compiled 23 examples of rocks of known ages and recorded when scientists actually watched them form. The results? The rocks were “producing excessively old K-Ar ‘ages'” when tested. In 23 of 23 cases, these radioisotope “ages” rose orders of magnitude above the rock’s actual ages. The culprit? Contrary to assumption, extra argon had entered the hardening volcanic rocks from sources other than radioisotope decay, greatly skewing the rock’s apparent isotope ages.

First, that’s a partial bait-and-switch between argon-argon and potassium-argon (K-Ar) dating.

Second, imagine you have a piece of paper, 100 micrometres (0.1 milimetres, or 0.004 inches) thick. You also have a wooden metre ruler – if you’re not familiar with the concept, this is a metre-long stick with distance markings down to the precision of 1 millimetre. You pass the paper and the ruler to a friend, and ask her to measure the paper’s thickness, which you already know. Assuming that she doesn’t cheat, she will get the wrong answer as the instrument just isn’t designed to measure things at the level of precision needed to get a reliable result in this case. But would you use this to cast doubt upon the reliability of the ruler when used properly? Would you declare on this basis a measurement of, say, 46 centimetres? That would be silly. Doubly silly would be for a third student to argue with his teacher that of course his measurements are inaccurate, metre rulers can’t measure anything! Unless, it seems, that teacher is Mr Thomas.

Argon dating is similar: it just can’t be used for dating any rock that scientists might have “watched form.” At No Answers in Genesis Kevin Henke provides a more thorough refutation, but the salient point about excess argon at the beginning is that as time passes it would become swamped by argon produced from the decay mechanism by which the rock is being dated and rendered irrelevant. Dating a young rock will get an incorrect result, but in older rocks this source of error vanishes.

So, absolute ages obtained from argon-related dating techniques should not be trusted. And they give different results than a separate uranium-based radioisotope “age” of only 33.2+1.5 million years that was found for the same region within the Kishenehn. Is the formation’s age 46 or 33 million, some figure in between, or none of these?

It would be wise to mention now that the Kishenehn formation as a whole is more than just the layer from which the mosquito fossil comes, and so many dates for different parts can be found. The 33.12 Mya figure comes from a 1996 paper by Kurt N. Constenius, which includes our previous numbers and says:

Single-crystal laser fusion 40Ar/39Ar dating of 12 biotite grains from this tephra resulted in an age of 46.2± 0.4 Ma (R. C. Walter, written commun., 1990). Fission-track analysis of 7 zircon grains from this tephra yielded a date of 43.5±4.9 Ma (C. W. Naeser, written commun., 1990; revised from 33.2±1.5 Ma reported by Constenius et al., 1989).

In short, Constenius himself had previously reported a figure of 33.2±1.5 Ma but based on a communication with a C. W. Naeser in 1990 he is revising this up to 43.5±4.9 Ma as seen previously. Why this was so is not clear, but Constenius seems confident that his previous result was wrong. Thomas is not, because he doesn’t realise that some numbers are more reliable than others and seems to think that any will do.

Moving on from argon dating, Thomas puts the horse behind the cart:

To narrow down the answer, secular scientists typically correlate the rock layer’s fossils to a geologic chart with numbered ages printed alongside fossil descriptions.

For example, one paper reported, “It [the Kishenehn Formation] contains a fauna of nonmarine mollusks and mammals, the latter permitting an age assignment to the Early Oligocene or Late Eocene”—ages that fall within an assumed 28- to 38-million-year range.

That quote is from a 1964 book, much older than our papers with actual radiometric dates. Lacking a more precise dating method you can use fossils to broadly pin down the age of a rock layer, but you don’t use fossils to more precisely calculate the date than you know from radiometric methods – Thomas has this completely backwards, and he must know it.** He says:

In another technical report, petroleum geologist Patrick Monahan wrote, “The Kishenehn Formation has a diverse fauna and flora that suggests a range of ages between late Eocene and early Miocene. However, a fission track age of 33.2+1.5Ma in the lower member in the Kishenehn Basin, and a K/Ar date of 29.9+5.3Ma in similar strata in a nearby basin indicate that the lower part of the Kishenehn is early to middle Oligocene.”

Is the formation from Eocene, Miocene, Oligocene, or none of these?

He quotes page 24 of a paper published in 2000, which cites Constenius’ earlier 33.2 Ma date but appears to have missed his later correction above. Thomas is digging up every potential date for the Kishenehn formation and its constituent layers, aiming to spread enough confusion that he can deny that the actual date can be determined. But that seems to be all he could find.

Now, Thomas sometimes ascribes to the view that different layers of strata correspond not to different times, but to different “ecosystems” that were inundated in the Flood and deposited on top of each other:

The idea that rock layers represented eras, like those named Eocene and Oligocene, actually surfaced hundreds of years ago when European naturalists decided different kinds of fossils are found in different rock layers because those creatures lived and died in separate ancient times. It became established dogma.

But other models might actually fare better. For example, what if each rock layer represents animals and plants from a particular ecosystem that was inundated and deposited by a tsunami-like wave, resulting in strata that show unique ecosystems, not separate times?

This is not borne out by the evidence: birds live everywhere today, for example, so how and why were they excluded from the first ecosystems to be inundated? The pattern makes no sense in this “model.”

Believing that he has sowed enough confusion, Thomas changes tack and heads towards the conclusion of his lengthy article. He makes two attempts to summarise his points, which you can read for yourself, before saying:

If the Kishenehn Formation is younger than 46, 43, 33, or 30 million years—if it is in fact only thousands of years old—then that would explain why it still contains abundant biodegradable oil and fresh, red blood protein remnants.

The “why is there still oil” question was raised last time, but again without any evidence that it should have degraded in the intervening period. And, once again, no “fresh, red blood protein remnants,” so this whole exercise was pretty pointless.

What have we learnt? First, we may have another math problem – in 2012 Thomas posted and then retracted an article on Enceladus that rested upon a calculation that overestimated the rate of mass loss of the moon by three orders of magnitude. Returning to the issue with the number-line above, he said that the full error range was “38.6 to 46.6 million years,” which is incorrect (the top of the range should be 48.4). I suspect, but don’t know, that his mistake here is responsible for the apparent mathematics illiteracy of that paragraph already explained. Failing that, it sounds like you could scam easy marks of the man by sowing confusion about the reliability of any measuring tool and then substituting a number you made up.

Secondly, the ICR is still pushing their old studies of recently-formed volcanic rock that supposedly challenge argon dating techniques. I’ve been waiting to use that paper-thickness analogy for a while, and there’s also a flip-side that comes into play with carbon dating as well.

Thirdly, Thomas still believes there’s something to his ecosystems idea, which I hadn’t seen for a while. I had thought that it had been overtaken by a hypothesis involving floating rafts of detritus which supposedly sank from the outside in, but I haven’t seen that one recently either.

And finally, Thomas is still trying to suggest that intact haemoglobin was found in this fossil, which is not the case. But that’s not so unexpected.

---

*Homo rudolfensis was a victim of the Dmanisi lump – but that doesn’t mean that it didn’t exist, merely that it should now be considered a subspecies of H. erectus, and referred to as H. erectus rudolfensis.

**He does know it – in one of his summary paragraphs he actually gets it slightly closer to correct:

The age of the Kishenehn came from matching its fossils with those listed and dated on the geologic chart. Radioisotope ages were then hand-picked to match the fossil-related age range from the chart.

Of course, the radiometric dates were not “hand picked,” but at least he realises this time that the less precise ages based on fossils came first.