Humans Nearly Went Extinct 1.2M Years Ago 356
Hugh Pickens writes "Scientific American has a story on researchers from the University of Utah who have calculated that 1.2 million years ago, at a time when our ancestors Homo erectus, H. ergaster, and archaic H. sapiens were spreading through Africa, Europe, and Asia, there were probably only about 18,500 individuals capable of breeding in all these species together (PNAS paper here). Pre-humans were an endangered species with a smaller population than today's gorillas and chimpanzees. Researchers scanned two completely sequenced modern human genomes for a type of mobile element called Alu sequences, then compared the nucleotides in these old regions with the overall diversity in the two genomes to estimate differences in effective population size, and thus genetic diversity between modern and early humans. Human geneticist Lynn Jorde says that the diminished genetic diversity one million years ago suggests human ancestors experienced a catastrophic event at that time as devastating as the Toba super-volcano in Indonesia that triggered a nuclear winter and is thought to have nearly annihilated humans 70,000 years ago."
Re:"Nuclear" Winter (Score:5, Informative)
Summary is wrong (Score:5, Informative)
The 18500 people quoted is not the number of people capable of breeding, but the "effective population", an abstract measure of genetic diversity in a species. According to TFA, the effective population of modern humanity is about 10000, and the argument in the article is that this much lower diversity indicates that a lot of genetic material must have been lost in a near-extinction event.
Re:Toba volcano ? Nuclear winter ? (Score:2, Informative)
Re:The new dogma of genetics (Score:5, Informative)
But why should we assume a uniform rate over time, when evolutionary theory says that genetic differentiation happens in leaps and bounds?
See, here is your problem, you're assuming evolutionary theory is correct to begin with.
Indeed, much of science is based on a giant leap of faith in linear regression; physicists, chemists, doctors, engineers, all use linear regression without questioning its assumptions.
No, they use linear regression and then test to prove it's a reasonable assumption.
Re:Nuclear Volcano? (Score:5, Informative)
What happened to those? Sounds like an excellent power source...
http://en.wikipedia.org/wiki/Geothermal_gradient [wikipedia.org]
"The Earth's internal heat comes from a combination of residual heat from planetary accretion (about 20%) and heat produced through radioactive decay (80%)"
In a sense, those "green geothermal" power plants are really nuclear power plants.
Re:"Nuclear" Winter (Score:2, Informative)
Re:The new dogma of genetics (Score:1, Informative)
Science seeks to explain, ie. to make complex things plain and amenable to human understanding, which is by definition a reductionistic activity. That's hardly a new insight. Your attempt to blame science for the simplicity of its explanations betrays a fundamental misunderstanding of what science can do. Science that is so impressed by "complexity" that it shrinks back from it is obscurantism.
Oh yes - "I used to be an atheist" is a complete non-sequitur in my opinion, especially if the opinions that accompany the statement are an embarrassment to any reasonable theory of science. Get your PhD in statistics if you like. But if your genetics courses were so bad, what makes you so sure that your stats curriculum isn't equally flawed? And have you ever actually talked with your engineering friends? I happen to be an engineer, and have used linear regression, being painfully aware of its assumptions and limitations. There's absolutely no "leap of faith" here. It's acquiring a mental toolset and learning how to use it appropriately.
Re:This means ... (Score:5, Informative)
You can find just about any belief "in some societies" if you look hard enough. The reality though is that in the English language the words brother and sister have a specific meaning: persons who share at least 1 biological parent. Some relatives from millions of years ago don't count.
Besides, from a biological standpoint, once you're to the genetic "distance" of only first cousins (1/8th DNA in common) the chance of birth defects drops off to the point of being completely fine. Indeed "in some cultures" marriages between cousins is very common (heck even in the US with all the attached stigma it's still perfectly legal for cousins to marry in most states). Thinking "we're all related man!" is only a problem from the standpoint of cultural taboo. Beyond the very immediate family it's not a problem.
Re:Nuclear Volcano? (Score:1, Informative)
The short answer is "yes". It's done all the time and it has been known since the early 1900s. In fact, it was radioactivity in the Earth that solved a major puzzle at that time: why the Earth was still so hot inside. Even after a few tens of millions of years it should have cooled down from a molten state much more by now if the only source of heat was what was left over from its formation. Once radioactivity was discovered, people realized that it would keep the Earth "hot" for much, much longer.
Anyway, it's fairly simple to take a sample of ordinary rock -- say, a nice piece of granite -- and measure the decay rate of the contained uranium, thorium, and potassium, which are the 3 big contributors to radioactivity, and therefore figure out how much heat is derived from it. There is a nice correlation between regional geothermal gradients and the radioactivity of crustal rocks of different compositions (e.g., areas with higher concentrations of radioactive elements in the rocks tend to have higher geothermal gradients), although you have to be careful about effects from rock conductivity variation, thinning of the crust, or underlying mantle plumes. Anyway, as an example, the geothermal gradient for continental crust is higher than would be expected for its thickness, on average, than for oceanic crust of the same age because continental crust contains much more potassium, uranium, and thorium. That difference is due to the chemical differentiation [wikipedia.org] of the continental crust (it is more "felsic" -- enriched in Ca, K, U, Th, etc., whereas oceanic crust is "mafic"). In other words, although thick continental crust blankets the underlying mantle better than the thin oceanic crust and should yield a much lower geothermal gradient because of that effect, more heat is produced by "in place" radioactivity in continental crust. By contrast, most of the heat flow seen in the comparatively thin oceanic crust is from heat escaping through it from underneath, and relatively little is produced in place. Continental crust still has a lower geothermal gradient than oceanic crust, but the difference would be much greater if not for the measured radioactivity.
While it isn't possible to measure the radioactivity of rocks from very deep in the mantle, and there are big questions about how radioactive the metallic core of the Earth might be because we don't have samples of it (check out the georeactor hypothesis), the numbers do add up pretty well for the samples we have that are closer to the surface.
Re:say that to the tasmanian wolf (Score:2, Informative)
Re:This means ... (Score:5, Informative)
First cousins have a common ancestor two generations back. So 53rd cousins should have a common ancestor 54 generations back.
54 generations ago, you had (theoretically) 2^54 ancestors (~180,000 trillion). Which means that statistically speaking, every human alive ~1200 years ago was about 200,000,000 of your ancestors.
In other words, such a number is pretty much meaningless....
Re:"Nuclear" Winter (Score:4, Informative)
Of course, when the range between eruptions is 660K to 800K, the low end of that would still make it another 20K years before the next eruption, or roughly twice the entire length of the history of human civilization. "About due" in geological time is very different from most people's view of "about due".
Yes, I'm aware the eruption could come earlier than previously observed, but it's not really worth worrying about events with astronomical odds that you can do nothing about now is it?
Re:"Nuclear" Winter (Score:2, Informative)
humor isn't so unique (Score:1, Informative)
non-human animals have humor. they play jokes on each other in nature and sometimes laugh. this is not intended to be a funny comment.
Re:Pfft... (Score:5, Informative)
We also have written evidence that Frodo set forth from the Shire in order to destroy the One Ring before it fell into the hands of Sauron. But so what?