Evolution as movement within a ‘fitness landscape’

A third noteworthy way of conceptualising the process of natural selection is in terms of movement within a ‘fitness landscape.’ The concept of the fitness landscape was first introduced by Sewall Wright in the 1930s and has proven a useful way to visualise how the addition of ‘parameters’ will influence the possible paths of evolution. This in turn can help us apply the abstract idea of evolution to actual biology. The landscape is typically pictured as a mountain range with several peaks of varying height, and valleys in-between. The height at any point on the landscape corresponds to its fitness value; i.e. the higher the point, the greater the fitness of an organism that occupies that spot.

The algorithmic conception makes clear how an organism undergoing Darwinian natural selection will tend to increase its fitness value. When viewed in terms of fitness landscapes it explains how the organism will move ‘up hill.’ As we saw above, an over-simplistic view of natural selection as an algorithm results in the fitness tending to infinity, but when contextualised on a fitness landscape, the possible fitness values are constrained to just those in the terrain; i.e. it can reach the top of a mountain or plateau, but can go no further. Richard Dawkins used the model of fitness landscapes to great effect in his 1996 book Climbing Mount Improbable.

While this conceptual tool adds needed complexity to a model of evolution, it has several limitations. For example, it gives the false impression that the landscape of possible adaptations is fixed. A more realistic representation of the relationship between organisms, environment and fitness would show the landscape changing as a result of the movement (i.e. adaptation) of organisms within it. It also does not make sufficiently clear the influence that adaptations in one organism will have on the fitness landscape of its peers. As Dennett says: “there is a tight interaction between the shape of the fitness landscape and the population that occupies it, creating a series of feedback loops ...The landscape is constantly shifting under your feet.”

It is also tempting to assume that organisms that reach higher up the tallest peaks are ‘better’ than those lower down or on shorter peaks. Like all models and metaphors fitness landscapes have their limitations. The height of a point on the landscape does not indicate a score of ‘goodness’ or complexity, but solely the fitness of that organism to that local environment.

Importantly, it must be acknowledged that coming up with an objective value for fitness is harder than it may seem. Working from a traditionally Darwinian perspective, we might propose that fitness could be measured in terms of number of progeny produced over unit time, but this is clearly inadequate since it does not consider elephants to be very fit at all. Other possible measures of fitness include: the complexity of individual organisms, the ability to survive in multiple habitats, or to evolve quickly to changing environments, or total biomass, or the ability to respond intelligently to complex problems. Gould is quick to point out that for three out of five of these measures bacteria are superior to humans.

A detailed model of the overall fitness of organisms would probably require several different landscapes, each representing a different aspect of adaptations that influences overall fitness.

 Printer-friendly | Feedback | Contributed by: Adrian Wyard