The speed of evolution

22 10 2008

The theory of evolution is greater than it looks. It is not just clever. It is not just useful. Its biggest value is as a nail in the coffin of some very destructive ideas. Not just the idea that Europeans are superior to Africans, or the idea that humans are superior to animals, but the idea that we all have some divine purpose – and therewith, the whole idea of good and evil.

The fascinating story of how the the tide of evidence has led to the unravelling of religious explanations for the world is, however, not what I wish to ponder here. No, I would like to ponder an area of evolutionary theory that still holds some uncertainty, some mystery.

Relax, I am not trying to ‘break’ or disprove evolution. I am fairly confident it it largely right, but I still think there are questions about it speed.

The problem…

Anyone who has read on the subject understands the pure cunning of natural selection. Basically put, any replicating ‘creature’, that produces slight mutations in its offspring, will produce some offspring that are better than itself – better at competing for resource, better at surviving. Of course many mutations (indeed perhaps most mutations) may produce ‘worse’ offspring, but if the better offspring survive proportionally more, there will be a generational improvement.

This is the same phenomena that allows us to breed better race-horses, beef-cattle or strawberries.

Now, we can see the effects of selection very quickly in a petri dish of bugs, or perhaps in viruses in the human population, but the evolution of large mammals is a slow affair, not easily observed, and it took the discovery of ‘missing links’ to confirm the theory that we had indeed evolved from primate stock.

I personally have not read widely on evolution, I have simply spent lots of time thinking about it, and also spent some brain cells on pedantic calculations and computer simulations.

What comes up, again and again in the simulations is the question of speed.


Yes, speed. How fast do we evolve, and have we had enough time to do it?


There are two ways to tackle the question of evolutionary speed. One the one hand, you could say: we have only had, say, 5 billion years, to evolve from the basic elements, so we must have evolved fast enough. The calculations must simply be made to fit the data.

Some (not me) have however said, hang on, calculations show that we haven’t had the time to evolve, so the theory must have some massive fault.

The latter argument betrays a misunderstanding of evolution. They assume that as evolutionists claim evolution is ‘true’ and ‘right’, that their models must be right. But if their models suggest we needed 100 billion years to evolve that will prove that evolution is too slow and some other agency is required to square the circle.

However, just because evolution is fairly certain to be right, that doesn’t mean the models are simple, and I hope to give some insight into the challenge that I came across in my own amatuer attempts at the challenge.

Factors that throttle evolutionary change… 

Let’s look at the things that effect the speed of evolution.

  1. the generation gap (time between generations)
  2. the strength of the mutation.
  3. the selection pressures (multiple)
  4. the male/female requirement (and its surprising turbo function)

The first one is obvious – the more generations you get through each year/millennium, the greater potential for evolutionary change.

Mutation strength is more interesting. You could have multiple errors in a DNA sequence, the more errors the stronger the mutation. However, some errors in the DNA may have no particular effect, while other errors could be catastrophic, so that matters too. To keep things simple, lets just focus on the ‘strength’ of the inter-generational change.

If the mutations were very small and subtle, this, I would predict, would slow evolution down. However, if the mutations are too large (remember they are random), they are less likely ‘to be compatible with life’. However, I suspect because they are random, they will come in all shapes and sizes, ranging from untraceably small – to very fatal (resulting in early miscarriage).

However, we have seen in the fossil record evidence that evolution speeds up and slows down. The statistics of mutation ought not to change like that, so there must be more to it.

Selection pressure is the next, and even more interesting, factor.

People have questioned why the world doesn’t have living examples of ‘the missing link’. They must have been ‘viable’ in their own right, so why didn’t they survive?

Some thought on the subject (as well as my own simulations) show that this is not surprising. The speciation event (when one species splits into two) is usually the result of a population becoming subjected to a varying selection pressure (usually geographical). If the population is well mixed, its keeps together, but a mountain range or body or water can reduce the interaction enough to allow the different ‘random walks’ to optimise the two populations for their environments. Allowed to proceed for any length of time, you land up with two separate species, with nothing between. Once separated, they cannot ‘rejoin’ even if they mix once more, so the divergence will continue. Some have argued that ‘spurts’ in evolution augment this speciation process.

So why do these spurts happen? This is most likely selection pressure, but how does it work? Well, putting it simply, the more the environment changes, the more the creatures will need to change to survive. In a static environment, creatures will evolve to suit, but following the law of diminishing returns, once happy, would have no driving force to change any more.

However, you could well argue, that while a quickly changing environment allows faster evolution, it is simply taking its foot off the brake, something else is really controlling the maximum speed of evolution.

What do I mean by maximum speed? Well ask how fast would an environment need to change, such that evolution could not keep up – then you have reached its maximum speed.

There is an example that some folks think is an example of an environment that promoted fast evolution. Theorists have suggested that early man was often the victim of famine, and often needed to move around, which resulted in accelerated evolution. The stronger penalty on the weak, the higher reward for strength and wisdom, meant that potentially positive mutations, that may be lost in stable ‘easy’ environments, were more effective in this environment, thus accelerating change.

There is much written about selection, by people much smarter than myself, so I will not elaborate any more on it. I would simply summarise, that in the course of billions of years, some environments would allow fast evolution, while others would stagnate, and the average speed, while hard to quantify, would not be consistently high enough to be the key to the sort of evolutionary speed we need to gain so many evolved features is so few generations.

So at this stage, I would like to point out the most marvellous thing about evolution, which I think can multiply the effects of natural selection.

The two-sex system and its accelerating effect…

As I grew up, I sometimes wondered why we needed two sexes. Why not just allow all creatures to have offspring, add in a little mutation, and hey presto, it should work. 

This was the form of my very first simulations (some 17 years ago now!). It showed me fairly rapid evolution and increasing fitness, so all looked good. If you set the bifurcation ratio to 2, give each child a random ‘fitness’ and then set the chance of reproduction be proportional to fitness (a very simple algorithm), the population did get fitter. However, when I compared it with the standard model, where you have two sexes, two differences in macro behaviour showed up.

Firstly, you have to add an extra ‘selection’ criteria. It is not just about surviving to reproductive age, it is now about surviving AND finding a mate. And you can forget about monogamy, so a very fit (for sexual selection) male could fertilise several females, at the expense of less fit males. This effect can (in my model) greatly accelerate change. All you need to do to get fast change, is ensure that partners can identify fitness accurately. 

The second interesting effect from having two sexes (and what I found out from my model), is that good genes can spread through the population, which does not happen in the one-sex model. This spread of good genes means that one part of a population could be learning how to deal with sunburn, while another is learning to deal with sickle-cell anemia, and their solutions can be shared by all.

Now we are talking. This, if I have my thinking right, would be a serious turbo-boost for evolution, allowing it to evolve lots of traits in parallel, whereas the single sex model would have to work on one feature at a time – first build an eye, then build a digestion tract, then build a good sense of humour… This buys evolution a lot of time. This makes it far more likely that 5 billion years is enough time to make all the amazing variety we see today.

So I would therefore argue that this makes it more likely we had time to go from the first two-sex replicator to the world of wasps, earwigs and herpes.

That begs my next question: how did the first two-sex replicator come about? I think the computer modelling may be beyond me, but I hold out hope for my children!