[Lewis Thomson (OP)]

Donald has presented this conundrum to The Naked Scientists,

"It seems evolutionary change though controlled by a physical molecule (DNA/RNA) rarely reversed it's direction, or undoes/reversed a change. Why don't organisms revert to prior, highly successful forms when the stress that made them change is eliminated?  Like nothing has even tried to be another dinosaur."

What do you think? Discuss in the comments below...

[alancalverd]

Stress doesn't make the change but selects for those natural variants that can tolerate it. Given the enormous range of possible variants and the quasicontinuous nature of evolution, the probability of a random mutation exactly matching a previous version, multiplied by the likelihood that the environment had itself reverted precisely (i.e. no new predators or diseases have persisted)  is extremely small.   

That said, it may indeed be possible to deliberately re-engineer an extinct species if we know enough about its critical genome sequences, and several people are working on recreating the woolly mammoth.

[evan_au]

Ecosystems can revert back when conditions change: In the exclusion zone around the abandoned Chernobyl reactor (in Ukraine), many native animals have returned.
- But this only happens if those species (and their genes) haven't been driven into extinction.

A similar thing happens with individual species: There are different variants of genes, and these gene variants can come back - provided they haven't been driven to extinction.
- An often-quoted (but somewhat controversial) example is the peppered moth, which developed a dark color in polluted city areas during the industrial revolution, but the original light color persisted in less-polluted rural areas. As pollution levels decreased in cities, the light colored variant spread back into previously polluted areas.
See: https://en.wikipedia.org/wiki/Peppered_moth#Evolution

If you take something like the dinosaurs, so many genetic variants went extinct during the meteor impact that these variants can't come back. Another group of animals (the mammals) eventually filled the newly vacated ecological niche, and now those dinosaur variants can't come back, because there is no ecological niche for them (except in Jurassic Park).

[chiralSPO]

There isn't any evidence that DNA changes to adapt organisms to environmental changes (just talking genetics here, not epigenetics). It is generally accepted that mutations are essentially random. While most single mutations have little to no effects, as the DNA changes, some changes lead to fatal errors, some to potentially harmful outcomes, and some to potentially beneficial outcomes.

Here's the thing about going backwards though. Even organisms with very simple genetic codes have more than 150,000 base pairs (and most have several million or billion of them, but let's stick with the simple case). Each of these 150,000 units has four options. So the chances of even a single mutation being exactly undone by another random mutation is very small. Exactly undoing 10 single-point mutations would be astronomically small! One species reverting to a previous form is essentially impossible (ie wouldn't likely happen by chance at any point in the next few billion years before the sun swells up destroys all life on earth.

[hamdani yusuf]

It is generally accepted that mutations are essentially random.

I think there's something more to consider, as I posted in another thread.

https://phys.org/news/2022-01-evolutionary-theory-dna-mutations-random.html
Study challenges evolutionary theory that DNA mutations are random

Jan 12, 2022

Study challenges evolutionary theory that DNA mutations are random
by UC Davis

Study challenges evolutionary theory that DNA mutations are random
Studying the genome of thale cress, a small flowering weed, led to a new understanding about DNA mutations. Credit: Pádraic Flood

A simple roadside weed may hold the key to understanding and predicting DNA mutation, according to new research from University of California, Davis, and the Max Planck Institute for Developmental Biology in Germany.


 
The findings, published January 12 in the journal Nature, radically change our understanding of evolution and could one day help researchers breed better crops or even help humans fight cancer.

Mutations occur when DNA is damaged and left unrepaired, creating a new variation. The scientists wanted to know if mutation was purely random or something deeper. What they found was unexpected.

"We always thought of mutation as basically random across the genome," said Grey Monroe, an assistant professor in the UC Davis Department of Plant Sciences who is lead author on the paper. "It turns out that mutation is very non-random and it's non-random in a way that benefits the plant. It's a totally new way of thinking about mutation."


Instead of randomness they found patches of the genome with low mutation rates. In those patches, they were surprised to discover an over-representation of essential genes, such as those involved in cell growth and gene expression.

"These are the really important regions of the genome," Monroe said. "The areas that are the most biologically important are the ones being protected from mutation."

The areas are also sensitive to the harmful effects of new mutations. "DNA damage repair seems therefore to be particularly effective in these regions," Weigel added.

Some genes are more important than others. Some genes in other locus might have evolved to protect or auto-correcting those essential genes from mutation. Survivor bias may also play a role in the study. Specimens with altered essential genes may just die early which skewed the result.

[alancalverd]

dinosaur variants can't come back, because there is no ecological niche for them (except in Jurassic Park).

There are plenty of tropical swamps that could support almost anything we recognise from the Cretaceous period, including several varieties of politician and government inspector.

[Deecart]

Ecosystems can revert back when conditions change

This is exact when only the ratio of species change, and this also apply to every allele (the particular expression of a gene).
Allele frequency within a population of a particular specie can grow up or down.
What can happen is that a particular allele is eliminated (by randomness and this is especially true when the number of individuals is low, or by selection).
In this case, of course, the genetic reversion is very unlikely.

But the "result" of the selection can appear again (and this is what we see whe we do some studies using fossils... because we generaly dont care about ADN but we study the anatomy of the living forms).
The environment can produce the same anatomy even when there is not any genetic link : Evolutionary convergence

Convergent evolution is the independent evolution of similar features in species of different periods or epochs in time. Convergent evolution creates analogous structures that have similar form or function but were not present in the last common ancestor of those groups. The cladistic term for the same phenomenon is homoplasy. The recurrent evolution of flight is a classic example, as flying insects, birds, pterosaurs, and bats have independently evolved the useful capacity of flight. Functionally similar features that have arisen through convergent evolution are analogous, whereas homologous structures or traits have a common origin but can have dissimilar functions. Bird, bat, and pterosaur wings are analogous structures, but their forelimbs are homologous, sharing an ancestral state despite serving different functions.

https://en.wikipedia.org/wiki/Convergent_evolution

[evan_au]

The longest experiment on cellular evolution has now passed 70,000 generations of E.Coli.
- One result that made news was the "evolution" of a gene that could metabolize citrate in an oxygenated environment.
- In fact, E.Coli have all the cellular machinery to metabolize citrate in both aerobic & anaerobic environments, but it is normally turned off if there is oxygen but no glucose. What occurred was a series of (probably 3) changes in gene regulation which allowed one population of E.Coli to metabolize the abundant citrate in their growth medium. The other 11 populations have not developed a set of mutations that allow this capability.

There are suggestions that there have been mutations in unused parts of the genome, that disabled those genes. So if these E.Coli were returned to a more natural environment, it is unlikely that they would fully regain the lost functions (barring horizontal gene transfer..).

So the "1-way traffic" of genetic evolution is a function of entropy interacting with natural selection.
https://en.wikipedia.org/wiki/E._coli_long-term_evolution_experiment#Evolution_of_aerobic_citrate_usage_in_one_population

There is a similar regulatory mutation in humans - mammal babies have the ability to digest lactose, an ability which is usually lost after weaning. However, independent changes in gene regulation in several human populations enable lactose metabolism even in adult humans. It is thought that these regulatory changes were beneficial after the development of agriculture, when a ready supply of milk from cows and goats became available.
https://en.wikipedia.org/wiki/Lactase_persistence

[Petrochemicals]

Because life is fairly blunt really, any evolution that occurs is most likely more complex and niche. To become less complex is difficult due to recources, a simpler life form will be able to evolve into the gap without the need to sustain surplus complex attributes. A highly complex organism will need to sustain itself  so will evolve or try to, to fit with its environment in its complex state. Likely it will fail and end up in an evolutionary culdesac.

[Deecart]

Specialisation and even hyperspecialisation do not mean evolutionary cul-de-sac.
Here an article about this :

Passion flowers with long nectar tubes depend entirely on the sword-billed hummingbird for pollination. However, as a new study by LMU researchers shows, the evolution of even such extreme specialization is by no means irreversible.
...
"This highly specialized mode of pollination is the result of a process of coevolution," says LMU's Professor Susanne Renner. "Such highly specialized adaptations need time to evolve, and this has led to the notion that their evolutionary trajectory is set in the direction of further refinement and can never be reversed." But, it turns out, evolution is no one-way street. By applying molecular phylogenetics and a so-called molecular clock, Renner and her colleagues Stefan Abrahamczyk (now at the University of Bonn) and Daniel Souto-Vilarós were able to show that the dependency of Tacsonia species on Ensifera ensifera for pollination has been lost several times over the course of a relatively brief period, geologically speaking.

Evolution can go into reverse

https://phys.org/news/2014-10-evolutionary-cul-de-sac.html