Sep 19 2016

Bacteria Evolving Resistance

Researchers at Harvard did a clever thing. They created a giant plate on which to grow bacteria, and included in the plate increasing concentrations of an antibiotic as you moved toward the center. They then plated bacteria on the outer edges and made a time-lapse video of the bacteria growing.

The end result was a video showing the evolution of progressive antibiotic resistance in the bacteria. You can actually see adaptive radiation, as the bacteria push up against the boundary to the next higher concentration of antibiotics, then multiple locations start to spawn new colonies spreading in the next zone.

The researcher made some more nuanced observations as well. For example, the bacterial grow slowed with new mutations, meaning they sacrificed something with the resistance mutation, but then they sped up again as they further evolved. Further, the most resistant bacteria were often not at the leading edge but were stuck behind less resistant bacteria.

The experiment resulted in an excellent visualization of evolution in process. Of course, this one piece of evidence does not “prove” something as complex and far ranging as the evolution of life on Earth. It is, however, a nice demonstration of evolution at work in a limited context.

Enter the Creationists

Creationists are evolution deniers. They deny, to varying degrees, the science of evolution. At the extreme end are those who believe in a recent creation and therefore deny evolution completely. But there are also those at the other end of the spectrum who acknowledge most of evolution but reserve one tiny slice that requires divine intervention. As long as they can demonstrate that there is one aspect of life that unguided evolution cannot explain, then they have proven God or whatever they imagine the “intelligent designer” to be.

Michael Behe occupies this latter end of the spectrum. In his recent book, The Edge of Evolution, he acknowledges common descent, a 4 billion year history of life on Earth. and even mutations and natural selection as the drivers of evolution.

But, he argues, some evolutionary developments were too unlikely to happen by chance and therefore required an intelligent designer to occasionally step in and make specific mutations or combinations of mutations happen by fiat. He is objectively wrong, and has been corrected numerous times by scientists, but that hasn’t stopped him.

Behe is now arguing that the recent Harvard study is not evidence for evolution, but is rather “devolution” because the acquired mutations do not create anything they just “break” existing enzymes.

Behe is wrong because there is no such thing as “devolution.” Evolution is simply heritable change, any change, and that change can create more complexity or more simplicity. Further, altering a protein does not “degrade” it – that notion is based on the false premise that there is a “correct” sequence of amino acids in any particular protein.

Evolution just makes proteins different. Proteins perform “better” or “worse” only in so much that they contribute to the survival and reproduction of the individual. If it is better for the survival of the organism for an enzyme to be slower, then the slower enzyme is better for that organism.

This is just one of the conceptual mistakes that Behe makes. Keep in mind that Behe has a narrative purpose to his motivated reasoning. He wants to arrive at the conclusion that random mutations and natural selection are not capable of creating life as we see it. Something else is required.

Here is his logic (as David Levin deconstructed very nicely here): creating complex structures that serve an adaptive purpose may require multiple simultaneous mutations. The probability of multiple simultaneous mutations is vanishingly small, so small that they are unlikely to occur at any significant rate over the span of life on Earth.

This is basically his irreducible complexity argument. It is wrong for two main reasons. The first is basically the lottery fallacy – considering the odds of John Smith winning the lottery by chance alone and concluding it could not have happened by chance. Rather, you should consider the odds that anyone would win the lottery. This is actually pretty good.

Behe looks at life on Earth and asks – what are the odds that this specific pathway or protein or whatever evolved by chance alone. He is failing to consider that there may have been billions of possible solutions or pathways down which that creature’s ancestors could have evolved. Species that failed to adapt either migrated to an environment in which they could survive, or they went extinct.

In other words, Behe should not be asking what the odds are that this bit of complexity evolved, but rather what are the odds that any complexity evolved. It is difficult to know the number of potential complexities that never evolved – that number may dwarf the odds of any one bit evolving.

Right there Behe’s entire premise is demolished, but he makes further mistakes. In his book he uses as a cornerstone of his argument the evolution of chloroquine resistance in the parasite that causes malaria. His big argument is that such resistance requires the simultaneous occurrence of two mutations. While this is very unlikely, the very short life cycle and great numbers of parasites means that it will happen occasionally by chance. Vertebrates, however, have a much slower life cycle and lower population numbers.

Therefore (and this is basically what the entire book is about) nothing more complex than chloroquine resistance could have evolved through Darwinian processes alone.

Casey Luskin, writing for the propaganda blog Evolution News and Views, proclaims that Behe has been proven right all along because a paper came out showing that chloroquine resistance does indeed require two mutations, although there are two possible forms of the second mutation (so he was only sort-of right). However, this misses the point entirely, as Luskin is famous for doing.

There are at least a couple of ways in which functions that require multiple mutations can occur through random processes. The first is when each mutation provides a small sequential advantage. This is what destroyed Behe’s examples of irreducible complexity, as simpler versions of structures can still function well enough to provide some survival advantage.

There is also the fact, although unrelated (as far as we know) to the current example of chloroquine resistance, that simpler structures could have served another purpose entirely and then were coopted to their current function. Simpler eyes still function as light receptors, and simpler wings could have served as displays, or for gliding or trapping prey.

The second way in which Behe’s logic fails that is directly related to the current example is that evolution involved more than natural selection. There is something called genetic drift – mutations happen and spread throughout populations without any selective pressure being involved at all. In fact, populations of parasites with one of the needed mutations have been identified. Then, within those subpopulations, just one more mutation is needed to confer resistance.┬áThis means that Behe was objectively wrong.

Therefore – you can have sequential mutation whenever one mutation confers any advantage, or simply through genetic drift.

There is, of course, also the now iconic example of Lenski’s bacteria that evolved over tens of thousands of generations to metabolize citrate as a food source. This required multiple sequential mutations that occurred by chance in some of the populations.

These are real world examples of multiple mutations occurring resulting in specific adaptive change. They disprove Behe’s thesis.


Evolutionary science is a robust discipline with many scientists around the world doing research and engaging with each other to figure out all the complexities and nuances of what is a very complex process that occurred over billions of years.

Creationists, even those who are trying to be scientifically respectable, like Behe, are simply not engaging with the community. They are making arguments that are logically and factually refuted, but they fail to adapt to valid criticism.

That is because they are not engaged in genuine science, but in motivated reasoning.


729 responses so far