Hacker News new | past | comments | ask | show | jobs | submit login
110M year-old nodosaur is the best-preserved fossil of its kind (2017) (nationalgeographic.com)
244 points by djsumdog 13 days ago | hide | past | web | favorite | 77 comments

Previously, with better pictures: https://news.ycombinator.com/item?id=14326913 (2017; 117 comments)

Plus with comments from the author of the article.

Ok, we'll change from https://www.earthlymission.com/dinosaur-mummy-science-discov... to https://www.nationalgeographic.com/magazine/2017/06/dinosaur.... Thanks!

The NG article is pay walled.

If there's a workaround, it's ok. Users usually post workarounds in the thread. There's at least one in this thread.

This is in the FAQ at https://news.ycombinator.com/newsfaq.html and there's more explanation here:



And from that link, here's a video of the miners lifting the fossil out of the mine... SPOILER ALERT

only to watch it break into countless pieces!! :(


Even though you said it was going to break before I watched it, it was still very difficult and surprising to see it happen. Especially if you watch it from the beginning with all the work they put into it.

I've seen this in person, and almost every year I visit the museum and have the same reaction. I stand there in wonder looking at it for about 30 minutes. The connection to the past world feels absolutely visceral. You can get very close to it, and it is never too busy around the museum as it is so large. If you ever get a chance to visit this part of Canada (Drumheller), and particularly this museum, you should go for it.

I second this! I visited the Tyrrel museum in Drumheller and the Dinosaur provincial park that's near Patricia as a young adult about 20 years ago and it is still imprinted in my memory.

I had seen dinosaur skeletons in museums before but it didn't compare. Also guided tours through the limited access of the park were amazing, almost surreal, with fossils of dinosaur bones just popping out of the ground now and then plus the chance of seeing active digs.

When we were kids we used to canoe down the Red Deer river in the badlands area, with a fellow from my parent's canoe club who was a retired geologist. (I grew up in central Alberta and my parents were pretty serious canoers, they usually did faster moving whitewater but once in a while we'd do the Red Deer just for us kids) My sister and I would have our minds blown by him as he explained every rock to us. I doubt you could do what we did now, we'd climb all over the hoodoos and find fossils all over (mostly bison bones, but still the odd fossil.) We never took them home, and I believe anything of import was either given to the museum or he told them about it.

Lots of neat little caves, bones all over. Very awesome place. Especially when you're a dinosaur obsessed 8 year old.

There was one time we were going down the river and there was a cow on the bank stuck in the mud up almost up to its neck. I remember the adults going to find the rancher and a lot of fuss and the cow was eventually saved.

That sounds like an excellent idea for a trip. I'd love to do that with my kid, it would awesome.

> it was an enormous four-legged herbivore protected by a spiky, plated armor. It weighed approximately 3,000 pounds.

> To give you an idea of how intact the mummified nodosaur is: it still weighs 2,500 pounds!

How does it weight less as a stone fossil than as an organic life form mostly made of water?

Most stone is 2-3 times as dense as water


> How does it weight less as a stone fossil than as an organic life form mostly made of water?

> Most stone is 2-3 times as dense as water

Presumably because the water was completely filling a volume, whereas the stone is very much not.

Because it wasn't fossilized; it's mostly original material.

As far as I can gather, that's not true and the soft tissue was in fact mineralized. Dinosaur "mummies" are fossils of mummies. This dinosaur 'mummy' is fairly unique insofar the soft tissue was mineralized without first being desiccated.

Chemical traces of things like pigmentation can remain in fossilized soft tissue, I suppose that counts as "original material", but this thing isn't made out of meat anymore.

Sadly I don't believe that's possible. It would be beyond fantastic if it were.

Wow, that's wild, almost like Black Swan wild. I was so predisposed to assume fossils are solid stone that I missed that. Still, living things are mostly water, and I doubt the mummified 2500 pounds is mostly water. Another response to my question suggests that just the outer shell is stone in the inside is empty. In that case, the mass is coming from the stone shell, and is not very analogous to the living animal's mass, which still makes the phrasing in the article a bit dubious.

Perhaps though it contains some original carbon that can be used to do some very old radiocarbon dating, which could improve geological strata based dating of other fossil finds.

Only half of it is there. The back half of the thing is gone.

> How does it weight less as a stone fossil than as an organic life

Because you are comparing different things. The fossil comprises only the anterior part of the animal. The tail and posterior legs are missing.

Why did this Canadian energy company (Suncor) react to this properly? Theories? Surely there'd be immense economic pressure to just keep digging.

(Am I being too cynical?)

Suncor was right in the midst of a big investigation into 500 ducks that died in a tailings pond. Might have played a role.


That sounds like a plausible reason. Timeline checks out.

There's a number of reasons I think, including: 1. The fossil itself is beautiful and outstanding

2. There is less pressure to produce right now given oversupply and shipping constraints in Alberta

3.Suncor cares more about its reputation than some of the other players in the tar sands

4. The Royal Tyrell Museum is well known to most Albertans, kids go there on school trips and it would likely seem like the obvious thing to do (stop work) when presented with such a find.

5. I have no idea if there is a finders fee, but that fossil is probably more valuable than anything that loader was processing all day.

6. It's a dinosaur - most folks find them pretty cool. :)

Re 5) I don't think they knew how valuable it was when uncovering it, the region is filled with fossils. When I visited Royal Tyrrell Museum in 2016 they said they have more fossils in the storage than they have time to process them.

2) It happened in 2011. The oil price was around $100/barrel back then. Was there a regional oversupply back then?

This is completely speculative, but Suncor was filthy rich in 2011, so a temporary stoppage wouldn't have hurt the bottom line too much. Also, because of the Royal Tyrrell Museum, dinosaur fossils have become a part of Alberta's identity. The miners would have talked about the find and the company would have faced bad publicity if it had acted differently. Oil sands companies don't really need any more bad publicity than they already get.

Now think of all the fossils and historical artifacts we're not hearing about because they just kept digging.

Obligatory question: DNA?

Not likely. According to the research I've seen, the half-life of DNA nucleotide bonds in fossilized samples is on the order of a few hundred thousand years. But that's the half-life of each bond, which means sequences of non-trivial length will become fragmented much more quickly. After 110 million years, it seems very unlikely that anything sequenceable still exists, even in trace amounts.


Could someone knowledgeable in pchem explain how this works? I'm guessing that half of the substance doesn't deterministically decay after a specific amount of time. I imagine that the decay follows some probability distribution, which should mean some portion of the substance decays much faster or slower, right? Does some of it never decay at all?

See link to previous discussion, posted above


Might help.

DNA doesn't have a "half-life," there's no consistent inverse exponential decay that continues in perpetuity. It's like saying what's the half-life of a steak? Degradation depends on conditions.

But if they're fossilized, those broken bonds would still be roughly in the same locations, right? So not useful for sequencing, but maybe the information can still be recovered by some future not-yet-possible means.

An interesting question might be, assuming this animal has some descendants alive today, has its original gene sequence been better preserved in that heritage than within its own body?

DNA half-life basically wouldn’t be able to survive this long, no matter how well preserved I believe.

Even if DNA itself can't, it seems possible that with advanced enough technology there might be hope of recovery of something other than DNA that could be used to figure out what the DNA was.

Imagine for example a machine that takes off 1 layer of atoms at a time, painstakingly charting them, and then another layer of software that figures out probabilistically whether the arrangement of atoms means that a decayed strand of DNA was here... and then probabilistically adds together the the billions of decayed shreds of DNA.

Maybe the relative positions of the base pairs are still probabilistically informative despite decaying and many of them breaking apart. I don't know. But it seems like there's a plausible way to try to extract data from fossilized DNA.

The half-life of DNA is ~521 years at 13.1°C. This dino is more than 100 million years old.

Scientists have been working very diligently to try to recover DNA from millions of years ago, but the reality is any DNA found has a very good chance of being from bacteria & other organisms from the more recent past.

Still there are (disputed) claims to have found DNA that's millions of years old:


Not clear to me how this meme that the half-life of DNA of about 500 years appeared. There is no concept of half-life for chemicals, there's one only for radioisotopes.

For this particular mummified dinosaur, if they found lots of somewhat preserved soft tissue, maybe hundreds of pounds, chances are that there could be trillions of DNA segments. Very likely no single gene will be unbroken, but with many fragments broken in different places the theoretical possibility to reconstitute the genetic code is there. You also don't start from zero knowledge. Humans and birds share about 65% of the genetic code [1], and dinosaurs are closer to birds than humans are.

[1] https://education.seattlepi.com/animals-share-human-dna-sequ...

>There is no concept of half-life for chemicals<



half life is a general term that is frequently used in biochemistry, enzyme kinetics, chemistry and nuclear physics

as a general term it is converse of doubling time and is simply a measure of stability or observed activity


Ok, there's a concept of biological half-life, including the famous 5 hours for caffeine. Living bodies are extraordinarily similar though, for example we all have about the same temperature of 36 deg Celsius. Chemicals out in the open, that's a completely different story.

To be more specific: [1] is a published article in Current Biology where they state they were able to fully sequence the genome of two mammoths, one of which was 44.8k years old. That would be 86 half-lives according to the meme of 521y half-life for DNA. 2^(-86) is roughly 10^(-26). By the logic of DNA half-life, reconstituting a 44.8k y.o. mammoth genome would be beyond utterly ridiculous. Still, here we are.

[1] https://www.cell.com/current-biology/fulltext/S0960-9822(15)...

Well, what half-life do you think that article implies? I took a shot at the calculation, and I got 1375 years.* That suggests to me that, yes, the popular figure may be off, but what difference does it make when you're talking about dinosaurs?

I mean it seems like you are implicitly saying that it's a huge exaggeration to claim 120,000 half-lives since the dinosaurs went extinct whereas it might be as little as 47,000. I don't think you're wrong. But raising 2 to either power overflows my calculator, so I don't see how it matters.

*my calculation is very imprecise and ignorant of biology, but I think orders of magnitude are good enough when talking about exponentials.

- the sample was 180 mg

- the number of cells in that size sample might be 8e10.[1]

- they had 17x coverage

- log2(17/8e10) is about -32

- 44,000 years / 32 = 1375

[1] I got a number for cells/kg by assuming a human weighs 180 lbs and has the number of cells mentioned in this reference: https://pubmed.ncbi.nlm.nih.gov/23829164/

I don't think it works quite like that. A DNA molecule (some people prefer to call it a "strand") has billions of nucleotides (or "DNA letters"). There is one DNA molecule per chromosome, and a few dozens chromosomes in every single cell. For all practical purposes, the same chromosome in various cells has the exact same copy of the DNA molecule. After the organism's death, each DNA molecule will probably break down in a matter of days, at most. From this point of view a half-life of 521 years is an astronomic number, and totally false. After a few years there will be exactly zero intact DNA strands. But that won't make a difference anyway, as we are not able to read a full strand in one read anyway.

Instead we read fragments and stitch them together. Like you see a "GATACA" and then a "TACACCG" and hypothesize that they were both parts of the longer string "GATACACCG". The longer the intersection string, the most likely the stitching is correct. In the mammoth case, they had fragments longer than 60 letters. Stitching together all these fragments, they were able to produce the full string (of billions of letters) 17 independent times. That gives them a high level of confidence that what they did is correct.

In the case of dinosaurs. Will we ever be able to read fragments of tens of letters and stitch them together? Maybe, maybe not. If the DNA breaks down so much that you won't be able to see even pairs of letters, then we are totally out of luck. If we see pairs of letters, we probably see triplets, quadruplets, etc. Can we stitch together a long string out of quadruplets? Most likely not, but if we have 10-letter strings, probably yes.

Are we going to have a contamination problem? If most of the DNA fragments come from bacteria rather than the original dinosaur, will we be doomed? Most likely we'll be able to come up algorithms that will eliminate bacterial DNA genes out of the reconstituted genome. We are quite smart.

Is Jurassic Park in the near future. Not before we are able to revive mammoths. And a mammoth is difficult to create for the simple reason that it takes a very long time to gestate a baby elephant. If each iteration in your experiment takes one year, you are in danger of shipwrecking your career.

Can we overcome these challenges? You bet. Anytime soon? My guess is that not during my lifetime, at least. But then Bill Gates states that we overestimate what we can do in one week but underestimate what we can do in a decade. So, you never know.

"I don't think it works quite like that"

I'm sure it doesn't, but we agree that the recovery of the sequence implies a half life greater than 500 years, right? I was trying to get at how much greater, whether double or 100x or 1,000,000x, etc.

"From this point of view a half-life of 521 years is an astronomic number, and totally false"

Please set aside this point of view, because it wasn't mine. If they got 17x coverage, that's not literally from 17 totally intact cells, of course I know that. But I was interpreting the recoverable DNA in aggregate as being what the half-life refers to, as it were, proportional to 17 cells worth. If that's within a couple orders of magnitude, it's good enough for my point.

If you think the half life could be millions of years, because I don't understand gene sequencing, then it would be really cool if you could do your own estimate and show me why your number is different.

I don't think the genome sequence coverage number can be used to infer anything about how fast DNA decays, or even how much DNA was available to the researchers. For example, from wikipedia [1], I found out that the African elephant's genome was sequenced in 2009; the details can be found in [2], where the sequencing coverage is listed at 7x. Why 7x and not 1000x, or more? Most likely there are some limiting factors, such as compute costs, and people are not chasing higher coverage numbers for their own sake.

>If you think the half life could be millions of years

I don't think the half life could be millions of years. I simply think half life is not a meaningful concept for DNA.

Books decay. What's the half-life of books? How many books from a hundred years ago are still around nowadays? Let's be generous and say 0.1%, that's a half-life of about 10 years. That means there should be no books older than 300 years, yet I have seen with my own eyes a Gutenberg bible, printed well before 1500.

[1] https://en.wikipedia.org/wiki/List_of_sequenced_animal_genom...

[2] http://genome.ucsc.edu/cgi-bin/hgGateway?org=Elephant&db=lox...

I don't understand why you insist that a number that has a large uncertainty cannot be meaningfully specified, when you yourself demonstrate putting constraints on it. That is giving it a meaning.

There are about 49 Gutenberg bibles known to remain of 158-180 copies, I read. log2(49/169) = -1.8; 570/1.8 = 316 years for half of them to disappear. It's very meaningful to say that a number is between 10 and 1000. It's not 10,000; it's not 100,000, or 1,000,000. There are maybe 10^82 atoms in the universe. That's about 2^272 so saying I am confident the half life of even the best preserved, most cherished books is less than 1000 years means that a 300,000 year old book could not exist. Because even if there were 10^82 of them, not one would be left. If you really insist you cannot quantify it, you're saying there's a chance! Aren't you?

With DNA, it's not that I know the half life is 1375 years precisely, or have hardly any biological knowledge, but the little knowledge I have combined with exponential rates of loss constrains it so I can't imagine the upper bound is above 10,000 years.

The thing I am trying to express is that even huge uncertainty and almost total ignorance give you far more to work with than "nothing".

What if the decay does not follow an exponential law? For example a power law?

Here's an interesting quote from the wikipedia article on ancient DNA [1]

"Researchers in 2016 measured chloroplast DNA in marine sediment cores, and found diatom DNA dating back to 1.4 million years. This DNA had a half-life significantly longer than previous research, of up to 15,000 years. Kirkpatrick's team also found that DNA only decayed along a half-life rate until about 100 thousand years, at which point it followed a slower, power-law decay rate"

[1] https://en.wikipedia.org/wiki/Ancient_DNA

I mean, you're looking at figure 3 and the description thereof, but that seems to be a relative amount of DNA. I'd consider figure 1 instead. It appears that the half life is about 200,000 years for site U1343. At that rate, you still wouldn't have any dino DNA left even if the whole universe was made of it.

> A DNA molecule (some people prefer to call it a "strand") has billions of nucleotides (or "DNA letters").

A DNA molecule consists of two strands.

RNA is single stranded.

When I first said the half-life of DNA is ~521 years at 13.1°C I was careful to include that temperature which made it clear that the decay of any organic material over time has more variables than merely time. Here's the article where that number comes from.


Mammoth DNA preserved in permafrost can last longer. Temperature is a variable. But dinosaurs lived in a very warm time, much warmer than the Ice Age the mammoths lived in so there's no possibility of nature somehow preserving Jurassic fossils in ice/permafrost.

While this thing is being referred to as "mummified" the fossil itself is a result of gradual processes that replaced the original form of the corpse with minerals. It's not dried meat in there. The find was a few years ago, so much has been written on the topic, and you should read what the discussions of experts on the topic had to say.

I'm not sure if you read this link before but I encourage you to - it includes a nice summary of the issues.


When you're looking at things this old the reality is any DNA found has a very good chance of being from bacteria & other organisms/contamination from the more recent than 100 million years ago past. Even if you come up with a really clever tool to find ancient DNA you have no guarantees of finding dino DNA nor any way to filter out the DNA from all the organisms of the past, virtually all unsequenced, that could be contaminants.

With advanced enough technology, we might train an ML model to generate a genome that morphologically approximates the dinosaur, simulating the trillions of protein foldings and chemical interactions involved in each trial, until we have a genome that produces the dinosaur, physically.

Because we have no certain model of the dinosaur, behaviorally, we can guess at that and ML our way to that, too.

And boom: you have your ersatz sim dino, and it only took like three Matrioshka brains.

That is my understanding too. Also they mention mineral deposits so it seems like this is still at least somewhat a fossil?

110 million years old feels like a shoot from the hip. How do they estimate the age of something like this?

Usually using radiometric dating. Carbon-14 has too short of a halflife (~5000 years) to be useful for fossils, but potassium-40 has a long enough halflife (~1.2 billion years) that it can be used to date minerals going back to the formation of the earths crust--it has even been used to estimate when the moon was formed (4-5 billion years ago)!

Radiometric dating is also used on the layers around the fossil. In this case of the hydrocarbon deposit [0].

[0] https://science.sciencemag.org/content/308/5726/1293

There's a comment from the article's author here: https://news.ycombinator.com/item?id=14327802

The data is thought to be pretty accurate (±5M years) and it's from radiometric dating of the oil sands.

Wow. It's really no surprise that people in the past imagined dragons! With no theories around fossilisation, biological/geographical eons, evolution etc., imagine finding something like this (even just an exoskeleton fragment!)

Wow this is amazing! So exciting to see stuff like this.

It looks like an ancestor of the denosaur, which seems to be a much more advanced/evolved take on the same general design.

Must had been a very hostile world if you had to be that armored just to eat plants.

I wonder if they recovered any vegetation samples from its digestive system

Saw this in person at the Tyrell museum. Pretty cool.

Umm don't you mean denosaur??

This is a solid NodeJS joke. Can't believe I got downvoted. Probably by a know-nothing wannaVC.

Agreed. Came here to make the same joke, which means it must be funny.


This article appears to be from 2017. There isn't any new information on the nodosaur specimen.

Yah, it's also a clumsy plagiarization of our article: https://allthatsinteresting.com/nodosaur-dinosaur-mummy

It looks like their site is another on the list that lifts our articles and recirculates them on social media.

Yours is just a regurgitation of the National Geographic feature, so I’m not sure what legs you have to stand on here. Ironic.

Which is itself a summary of the original NG article? https://www.nationalgeographic.com/magazine/2017/06/dinosaur...

@dang can we change the link to this one to give credit to the source? And add 2017 tag.

We've changed from https://www.earthlymission.com/dinosaur-mummy-science-discov... to the original source.

You will get a faster response if you send an email to hn@ycombinator.com They usually respond very fast.

Always a red flag when an article doesn't put a date of publication anywhere on the page.

> To give you an idea of how intact the mummified nodosaur is: it still weighs 2,500 pounds!

Why is that relevant? Tissue has been replaced by minerals which presumably are much denser than flesh.

Trollish usernames aren't allowed on HN because they basically troll every thread they post to. This one's admittedly only mildly trollish, but still.


We've banned this account for now, but if you want to email hn@ycombinator.com with a better username, we can rename it for you and unban it.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact