[alexvr]

Isn't this obvious to anyone who takes a little time to think about evolution and biodiversity? Yes, little changes caused by chance persist if they don't hinder the creature's ability to survive & reproduce. And these relatively benign mutations can accumulate until one family is very different from one that was once more obviously related. I don't see how this "surprising" and "provocative" proposal that chance plays a role in biodiversity is at all "controversial" or even new. What's more, the journalist seems to think that this "finding" means she can downplay natural selection, when in fact this article should really be nothing more than an emphasis of the role of chance in speciation since natural selection is basically predicated on the biodiversity caused by chance mutations. This really seems like publishing something 100 years after the first airplane was built to say: "Hear ye! Amazing new finding: In the absence of wings or power, an airplane simply falls due to gravity!" Maybe the actual research/paper is of more merit than the article suggests?

[colin_mccabe]

In the 1950s, nobody would have believed that "genetic drift" happened at all. The reason was because back then, the thinking was that any genetic change must be adaptive and caused by natural selection. The accumulation of "neutral" changes-- genetic drift-- was very controversial when it was first proposed, and wasn't really widely accepted until at least the 1980s or 1990s.

This main thrust of this paper seems to be the idea that genetic drift is the primary driver of speciation (the process of creating new species.) The paper also makes claims about a "speciation clock" that seems to tick every 2 million years, for many different species.

These two claims seem somewhat at odds. Presumably genetic drift will happen more rapidly in species like fruit flies or beetles that have short generation times. Genetic drift should be slow in species like whales since they take so long to reproduce, and drift can't happen except during reproduction. So if species are splitting off every 2 million years for both whales and fruit flies, that suggests genetic drift can't be the only factor, since it's 1000x faster (or more) in fruit flies as in whales. Anyway, I didn't read the paper that closely, so maybe I'm missing something.

In general, speciation is a complicated and poorly understood topic with a lot of controversies remaining. For example, is it more common for a species to split into two because of a physical obstruction (body of water, mountain range, etc.), or because different sub-populations start occupying different ecological niches due to slightly different environments, and gradually grow more distinct? (allopatric speciation versus parapatric speciation). It's hard to actually measure how much mixing there was between populations millions of years ago, so this is still controversial. And parapatric speciation gives more weight to natural selection so it used to be the preferred explanation. But this is far from a settled debate.

[mohawk]

Just look at those graphs of distributions for speciation times. Reading the description of the "speciation clock", i would have imagined a sharp distribution with a very tall peak at 2 million years. Yes, the mode is about 2 million years. But the distribution is veeery broad. Looking at the graphs, there is no reason to believe that whales and flies have similar speciation times. A very misleading description, unfortunately quite typical of the scientific news cycle.

[andrewflnr]

I think the interesting conclusion is that most speciation, emergence of populations that an no longer interbreed, is the result of neutral mutations, not ones that are naturally selected for. Adaptive mutations, I suppose, become universal before they split the species.