Fitness Landscapes
The first time I saw a fitness landscape cartoon (in Garrett Hardin's Man And Nature, 1969),
I knew it was giving me advice on how not to get stuck
over-adapted—hence overspecialized—on some local peak of fitness, when
whole mountain ranges of opportunity could be glimpsed in the distance,
but getting to them involved venturing "downhill" into regions of lower
fitness. I learned to distrust optimality.
People as well as animals can become overspecialized with chaotic competition, this optimum can be in comparison to competitors that can also be heading for collapse. Fitness can be a tautology if it only describes survivors as fit, they might survive because of chaotic or random influences. For example an R plague of rodents might only have a few survivors, this might require breeding as fast as possible while eating anything available to be on of these survivors. Fitness then here is to minimize losses in a negative sum game when hitting floors and ceilings. Random fitness can just be luck, a Ro herd of buffalo might be targeted by a Y pride of lions and the buffalo taken by chance. However to be fit the buffalo might move to an equilibrium of average characteristics so the war of attrition by predators leaves plenty of similar buffalo to survive. If they were very different then crucial genes might be lost to these predators and the species changes or collapses.
...
The man behind the genius of fitness landscapes was the founding theorist of population genetics, Sewell Wright (1889-1988). In 1932 he came up with the landscape as a way to visualize and explain how biological populations escape the potential trap of a local peak by imagining what might drive their evolutionary "path" downhill from the peak toward other possibilities. Consider these six diagrams of his :
This high rate of mutation can occur with Oy predators attacking R prey in the Roy animal kingdom, instead of being normal Ro prey through interbreeding the R prey become more diverse. This also happens with the Oy predators that feed on them, they also mutate so the results for both are highly chaotic between ceiling of high populations and floors of low numbers. When the environment changes some mutations hit the ceiling of their capability and starve or get eaten, this allows the R species or their Oy predators to innovate and counter innovate in the new situation. This is not evolution but R revolution and Oy counter revolution.
The bottom row explores how small populations respond to inbreeding by wandering ineffectively. The best mode of exploration Wright deemed the final diagram, showing how a species can divide into an array of races that interact with one another. That jostling crowd explores well, and it can respond to opportunity.
Fitness landscapes express so much so economically.
Ro animals are more cooperative as are Y predators, because the environment is less specialized in its Biv vegetation then generalists do better rather than those specializing in Oy-R foods. An animal can randomly eat different foods and the diet averages out ok, an Oy-R animal can specialize in certain foods allowing them to grow exponentially in numbers but sometimes also crash as that food varies between floors and ceilings in availability.
There's no better way, for example, to show the different modes of evolution of a remote oceanic island and a continental jungle. The jungle is dense and "rugged" with steep peaks and valleys, isolating countless species on their tiny peaks of high specialization. The island, with its few species, is like a rolling landscape of gentle hills with species casually wandering over them, evolving into a whole array of Darwin's finches, say. The island creatures and plants "lazily" become defenseless against invaders from the mainland.
You realize that for each species, its landscape consists almost entirely of other species, all of them busy evolving right back. That's co-evolution. We are all each other's fitness landscapes.
People as well as animals can become overspecialized with chaotic competition, this optimum can be in comparison to competitors that can also be heading for collapse. Fitness can be a tautology if it only describes survivors as fit, they might survive because of chaotic or random influences. For example an R plague of rodents might only have a few survivors, this might require breeding as fast as possible while eating anything available to be on of these survivors. Fitness then here is to minimize losses in a negative sum game when hitting floors and ceilings. Random fitness can just be luck, a Ro herd of buffalo might be targeted by a Y pride of lions and the buffalo taken by chance. However to be fit the buffalo might move to an equilibrium of average characteristics so the war of attrition by predators leaves plenty of similar buffalo to survive. If they were very different then crucial genes might be lost to these predators and the species changes or collapses.
...
The man behind the genius of fitness landscapes was the founding theorist of population genetics, Sewell Wright (1889-1988). In 1932 he came up with the landscape as a way to visualize and explain how biological populations escape the potential trap of a local peak by imagining what might drive their evolutionary "path" downhill from the peak toward other possibilities. Consider these six diagrams of his :
[Image credit: © Sewall Wright, The Role of Mutation, Inbreeding, Crossbreeding, and Selection in Evolution, Sixth International Congress of Genetics, Brooklyn, NY: Brooklyn Botanical Garden, 1932.]
The
first two illustrate how low selection pressure or a high rate of
mutation (which comes with small populations) can broaden the range of a
species whereas intense selection pressure or a low mutation rate can
severely limit a species to the very peak of local fitness. The third
diagram shows what happens when the landscape itself shifts, and the
population has to evolve to shift with it.This high rate of mutation can occur with Oy predators attacking R prey in the Roy animal kingdom, instead of being normal Ro prey through interbreeding the R prey become more diverse. This also happens with the Oy predators that feed on them, they also mutate so the results for both are highly chaotic between ceiling of high populations and floors of low numbers. When the environment changes some mutations hit the ceiling of their capability and starve or get eaten, this allows the R species or their Oy predators to innovate and counter innovate in the new situation. This is not evolution but R revolution and Oy counter revolution.
The bottom row explores how small populations respond to inbreeding by wandering ineffectively. The best mode of exploration Wright deemed the final diagram, showing how a species can divide into an array of races that interact with one another. That jostling crowd explores well, and it can respond to opportunity.
Fitness landscapes express so much so economically.
Ro animals are more cooperative as are Y predators, because the environment is less specialized in its Biv vegetation then generalists do better rather than those specializing in Oy-R foods. An animal can randomly eat different foods and the diet averages out ok, an Oy-R animal can specialize in certain foods allowing them to grow exponentially in numbers but sometimes also crash as that food varies between floors and ceilings in availability.
There's no better way, for example, to show the different modes of evolution of a remote oceanic island and a continental jungle. The jungle is dense and "rugged" with steep peaks and valleys, isolating countless species on their tiny peaks of high specialization. The island, with its few species, is like a rolling landscape of gentle hills with species casually wandering over them, evolving into a whole array of Darwin's finches, say. The island creatures and plants "lazily" become defenseless against invaders from the mainland.
You realize that for each species, its landscape consists almost entirely of other species, all of them busy evolving right back. That's co-evolution. We are all each other's fitness landscapes.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.