Theory of Speciation; History of Life

I read Gould's "Structure of Evolutionary Theory" and I was relieved to see that the there are others who are struck with the same problem with Darwin that I have. The thing that doesn't add up in "Origin of Species" is slow gradual evolution couldn't work during extinction events. More specifically, the period afterwards when the environment has radically changed and that all living things are somehow expected to slowly catch up. The life timescale is far shorter than the evolutionary timescale. If that were true, the species would become extinct.

Darwin did provide a good first order solution to the problem, but Gould took it to the next level. What Gould did was to separate the ideas of tiny amounts of change and the rate of change. He proposed that changes in a species could be rapid or slow, but were usually rapid. He retained Darwin's idea of tiny changes occurring from generation to generation and that these changes must accumulate (thus dispelling any theory about monstrous or over night change), but the rate of change from one species to the next could be rapid and was usually followed by a period of evolutionary stagnation.

Gould did offer a good second order solution to the problem, but he failed to link the change in environment to the change in the species. In other words, Gould spoke of changing species, but no reason why they change. Then it occurred to me. If a speciation is a response to changes in the environment (which we've established is possible), then there must be a dynamic feedback control system involved. What I am proposing is modelling the species as a standard dynamic system with feedback. What emerges is a model of speciation which is tightly linked to the environment. It is highly generalized in that it works for any type of selection including cultural and sexual, and also explains both gradual and punctuated evolution as a two specific forms of change along a gradation.

A standard system includes a plant or place where work takes place. There is usually some type of control mechanism which turns on or controls the output of the plant. A feedback mechanism must be installed so the control can monitor the output. The feedback is almost always negative. Negative feedback draws down the change over time. This causes the system to "lock in" on a specific value. If this were a furnace, the plant would be the house it heats, the feedback is the heat sensor, and the control is the thermostat. If it is too cold the heat sensor measures this and the thermostat knows to kick on the furnace. Once the house is heated (i.e. the control mechanism detects this via the feedback, or heat sensor), it shuts off the furnace. This feedback is a kind of dampening mechanism, like shock absorbers on a car. The first "shock" is large and the reaction to it is almost as large. Successive vibrations become smaller and smaller until they are completely eliminated. Every system has a kind of "speed" associated with it. That speed is how long it takes the output to cause a change in the output. In other words, how fast can the system react to a change in the environment. A "perfect" system is immune to disturbances in the environment. Remember this, it will be important later on when we talk about rate of change in species. For more information, see the Wikipedia article on Control Theory.

In our case, the plant is the individual animal. This system will measure the number of mutations in response to environmental change. The reference input will be 0 mutations. 0 mutations implies the animal is in a perfect environment and this will always the be reference as the "goal" of nature is to produce perfectly optimal offspring. The sensor measures the number of mutations, or how well the species is adapted to the environment. The control biology of the animal will use this to trigger changes in the species. This is problematic for the experiment because the change is considered to be completely random and truly independent. If this is true, then (historically) why is change always greatest during times of environmental upheaval? There are several possible solutions, a) Gould's internal constraints, b) that the environment can have direct control over the reproduction capabilities of a species, and c) that change is random and this experiment will have to be rethought in terms of the species survival and not an individual animal's mutation.

Figure 1. Feedback model with direct control

               [deviation (all types) from optimal environment]_____
                                                                   |
                                        (a)                        |(b)
[0 mutations]-->+[adder]-->[mutation trigger control]-->[sex]-->[animal]-->[# of mutations]
[ reference ]      -|                                                    |
                    |____________________[adaptedness sensor]____________|


Figure 2. Feedback model with no direct control
               [deviation (all types) from optimal environment]_____________________________________
                                                                                                   |
                                                                                                   |(c)
[0 mutations]-->+[adder]-->[mutation trigger control]-->[sex]-->[animal]-->[# of mutations]-->[survival]

In Figure 1 the environment has direct control over the animal either via the sensor (b) or as some type of internal constraint (a) on the control "circuitry" of the animal. In either case a change is triggered. In Gould's case the change is independent of the environment. I don't agree with this, but let's just say it's a possibility. The other is change as a response to the change in environment. What is absolutely key is the sex step in the sequence. In evolution, the change doesn't occur in the animal, it occurs in its offspring. This is extremely important. This turns the feedback mechanism into a very slow enterprise. This model also says the rate of adaptation (the time it takes to dampen the changes) depends on the initial number of mutations which is in turn dependent on how radical the change in environment.

In Figure 2 mutation is completely random and there is absolutely no link from the environment to the animal in terms of mutation. We see, as we see in both cases the survival of the species depends on the survival on the animal. The key difference of course is that Figure 2 is the hit and miss model and Figure 1 is the dynamic system model.

As we can see, if adaptedness falls, then the amount of mutation increases. If the amount of mutation increases, then by the next iteration, the level of adaptedness will have increased. With enough iterations (and remember each iteration is a new generation caused by births) the number of mutations falls to zero. This is where Gould's theory of punctuated equilibrium comes in. An extremely large change in the environment will trigger a huge change in the species initially, but will trail off and eventually fall to zero. The fossil record is not detailed enough to show this, but perhaps lab tests with mosquitos will reveal this. In the case of the fossil record of sea animals, I predict that the animals underwent a constant, but small change in the environment (probably caused by their own effect on the environment). This model does not account for drift, but drift still works here because the DNA that undergoes change is not coupled to the environment. If the model were altered to account for some weighted output, we would see that the drift mutations are of a very low weight and consequently do not trigger any changes in the species.

However, this model does not tell us how successful a species is and it does not explain the phenomenon of extinction. If the model were to take in to account internal constraints, and by this I mean the ability of a species to change, we could easily model the success, or lack thereof, of a species. The changes to the Figure 1 model illustrate this:

                              [internal constraint]  [environmental disturbance]
                                       |                           |
                                       |                           |
                           [mutation trigger control]-->[sex]-->[animal]-->[# of mutations]
                                                                    \->[DEATH]

With this modified model we see that too much pressure results in death of the species. Eventually all members of that species dies if there is not enough adaptation. Internal constraint is a factor in the amount of mutation that is available. Species that remain flexible in the face of change maintain higher longevity and a bigger population size.

This more refined, dynamic, precise method to the problem could completely model the speciation cycle of an animal, in abstract. It would be nice to see some research in this area to prove or disprove my theory. It would be nice to see this more fully expanded as I am sure my theory is incomplete or erred. I would like to see some mathematical models that would allow us to model not only the speciation of animals in the fossil record (to test and prove the theory) but also to gauge not only how many species there might have been in the history of life but also to be able to tell us more about past weather and environment of this planet. Most importantly we can use this information to see just what kind of environmental pressure living species can endure and what kind of impact we are really having on the environment. Something like this could change our current "after the fact" approach to something more proactive.