Why do some bad traits evolve, and good ones don't?

Question

If a trait would be advantageous to an organism then why hasn't it evolved yet?

Conversely, if a trait is not advantageous or mildly disadvantageous, why does it exist?

In other words why does evolution not make the organism more "perfect"?

---

This is a general question that would be applicable for any kind of trait. Please keep the answers precise and scientific.

Read this meta post for more information: Questions asking for evolutionary reasons

Answer 1

During the process of selection, individuals having disadvantageous traits are weeded out. If the selection pressure isn't strong enough then mildly disadvantageous traits will continue to persist in the population.

So the reasons for why a trait is not evolved even though it may be advantageous to the organism, are:

• There is no strong pressure against the individuals not having that trait. In other words lack of the trait is not strongly disadvantageous.
• The trait might have a tradeoff which essentially makes no change to the overall fitness.
• Not enough time has elapsed for an advantageous mutation to get fixed. This doesn't mean that the mutation had not happened yet. It means that the situation that rendered the mutation advantageous had arisen quite recently. Consider the example of a mutation that confers resistance against a disease. The mutation wouldn't be advantageous if there was no disease. When a population encounters the disease for the first time, then the mutation would gain advantage but it will take some time to establish itself in the population.
• The rate for that specific mutation is low and therefore it has not yet happened. Mutation rates are not uniform across the genome and certain regions acquire mutations faster than the others. Irrespective of that, if the overall mutation rate is low then it would take a lot of time for a mutation to arise and until then its effects cannot be seen.
• The specific trait is too genetically distant: it cannot be the result of a mutation in a single generation. It might, conceivably, develop after successive generations, each mutating farther, but if the intervening mutations are at too much of a disadvantage, they will not survive to reproduce and allow a new generation to mutate further away from the original population.
• The disadvantage from not having the trait normally arises only after the reproductive stage of the individual's lifecycle is mostly over. This is a special case of "no strong pressure", because evolution selects genes, not the organism. In other words the beneficial mutation does not alter the reproductive fitness.
• Koinophillia resulted in the trait being unattractive to females. Since most mutations are detrimental females don't want to mate with anyone with an obvious mutation, since there is a high chance it will be harmful to their child. Thus females instinctually find any obvious physical difference unattractive, even if it would have been beneficial. This tends to limit the rate or ability for physical differences to appear in a large & stable mating community.

Evolution is not a directed process and it does not actively try to look for an optimum. The fitness of an individual does not have any meaning in the absence of the selection pressure.

---

If you have a relevant addition then please feel free to edit this answer.

Answer 2

There's always the most obvious: Evolution is chance.

Some traits allow an individual to have a higher chance to produce offspring. That doesn't mean individuals with that trait have more offspring, not even on average unless the law of large numbers applies. A randomly mutated perfect squirrel could appear, and since it's only one, it gets run over by a car before it can reproduce and all perfect traits are lost.

Then there is a chance that a trait is beneficial for the individual, but doesn't have a significant effect on its chance of having offspring. For example, take human eyesight. People with really poor eyesight have a disadvantage. But most people have good enough eyesight until they are too old to have kids, so it doesn't matter to evolution if they fall off a cliff afterwards. It may even be advantageous if kids don't have to spend resources to care for their parents.

And a very nice effect is a swarm effect where you look at entire swarms, for example of fish. Take a swarm or community that reproduces primarily within itself, where for some reason only traits that are beneficial for the individual can get inherited. As soon as the swarm gets into competition with a superior swarm - where traits that benefited the swarm got inherited - the entirety of the former swarm may die.

There's also the effect of society. If during one decade a society thinks people with green hair are not attractive partners - even though it's better for camouflage - it sucks to have green hair during that decade. This doesn't only apply to humans. I remember reading that among some bird species the singing evolves and which songs are "hip"/attractive depends on social factors as well as genes. Just imagine what would have happened if a small number of humans would have started to grow bat wings that made them fully capable of flight, during the time of the Spanish Inquisition - rather than spreading the new gene they would have burned on a stake.

Let's not forget about the really cool effect of frequency dependent selection where a trait is only beneficial if not too many other individuals have it. These traits can be beneficial for an individual, or a species, or both, but only if not too many individuals carry the trait. A community of humans can benefit from super smart people who have limited control of their bodies and emotions, but if all humans in the community are like that the community has a problem.

Last but not least, there's the change in external conditions. If conditions change, many species will simply go extinct (e.g. Ice Age). Imagine all viruses disappear. If there are no viruses, a trait that makes the immune system more efficient against bacteria and cancer at the expense of having no defense against viruses, is clearly a beneficial trade. Then viruses come back and every species that adapted the "no defense against viruses" trait goes extinct from a viral infection. This can make it beneficial to not adapt new traits too quickly.

In summary: By accepting that evolution is a random process where individuals with more advantageous traits are simply somewhat more likely to have more offspring, and where entire communities or subspecies are slightly more likely to go extinct if the traits are less beneficial for the community, you get all of the above effects and many more. Keep in mind that individuals with advantageous traits are in no way guaranteed to have more offspring, and communities can be wiped out or flourish due to all kinds of incidents that completely ignore how advantageous or disadvantageous their traits are.

Answer 3

Because evolution is an effect, not a cause. That is, there's no "God of Evolution" out there deciding that this or that trait would be beneficial to a species, and deciding to add it. Evolution just works* on whatever random variations happen to come along.

*And as others have pointed out, it works statistically, not deterministically.

Answer 4

Below are the reasons I can think of. The list is not exhaustive and there are some conceptual overlaps.

• The trait seems advantageous but it is not, maybe due to its effect on another component of fitness (trade-off). It sounds to me to be the most likely explanation whenever you are wondering why a given species does not carry a given trait. In other words, the fitness landscape matter.

• The trait is not that easy to build it requires a series of mutations that are either neutral, quasi-neutral or deleterious.

• The mutation just hasn't happened yet (lag load).

• the mutant variant had appeared several times but got extinct because of genetic drift even though it was beneficial (assuming no fitness valley to cross).

• Putting the two previous points together, if only one mutation (one SNP) is needed to build this trait and if the trait is beneficial but quasi-neutral, then the expected time before fixation to occur is around $10^9$ generations. Here is why. Let $N$ be the population size and $\mu$ be the mutation rate. In a diploid population (but the end result is the same for the haploids), there are $2n\mu$ appearing at this locus each generation. As the mutation is quasi-neutral, its probability of fixation (calculated from Wright-Fisher equations or Moran birth-death process, or from Kimura's diffusion equation) is $\frac{1}{2}N$. Therefore the rate at which fixation occur is $2N\mu \frac{1}{2N} = \mu$. $\mu$ for an SNP in a species that has a large genome is around $10^{-9}$. The probability that fixation occurs in $t$ generations is given by the exponential distribution with parameter $\mu$ ad the expected value of this exponential distribution is $\frac{1}{10^{-9}} = 10^9$ generations. If there is one generation per year, that takes a lot of time! Of course, those calculations are a good approximation when the trait of interest is quasi-neutral. If the positive selection is very very strong on this trait to the point that we can completely neglect drift, then the expected time for the mutant to occur is $\frac{1}{2N\mu}$, which is $10^5$ for a population size of $N = 10000$ which is still quite a lot. In the meantime, the environment or the genetic background may change so that the trait is not advantageous anymore. The above calculations assumed that the trait appeared with only one mutation and that only one SNP can cause the trait to exist. One can be more general by summing exponential distribution (for multiple mutation scenarios) and increasing the mutation rate (for mutations at different loci can make the trait exist).

• the mutant variant had appeared several times but got extinct because of drift because the Ne in the genomic region is very low due to the act that there are many loci under selection in the surroundings (background selection).

• the trait is beneficial and occur sometime but either stay at low frequency or even regularly disappear due to a migration of individuals coming from an environment where this trait is strongly deleterious (dispersal load).

• Be careful to understand what evolution does and therefore to understand what advantageous mean. Taking kin selection/group selection/lineage selection apart to make things easier, an advantageous allele is the one who increases the fitness of its carrier. The effect of a given allele depends on the environment and on the genetic background in which it exists. A beneficial allele does not necessarily increase survival or increase the probability of the population to not get extinct. Think about sexual selection typically.

Answer 5

Evolution occurs by a change in gene frequencies, with gene frequencies potentially affected by four mechanisms (mutation, migration, drift, and selection).

---

The answer to the question Why does the seemingly advantageous trait X not evolve? could be:

1. The mutations for a trait have never occurred within a population, or such genes have never migrated into the population

2. The mutation(s) to cause such a trait may have once been present in a population but migration and/or drift and/or selection has since removed them from the population

3. Selection has removed the mutations causing the seemingly advantageous trait because the same (or tightly linked) genes have other deleterious effects by pleiotropy

---

Similarly, the answer to the question Why does the seemingly disadvantageous trait X evolve? could be any of the following:

1. Alternative genes have never been introduced to the population by mutation or migration

2. Mutation, drift, and migration have allowed the gene to fix over other more advantageous genes

3. Selection has spread the mutations causing the seemingly disadvantageous trait because the same (or tightly linked) genes have other beneficial effects by pleiotropy

---

• Note also that the trait may seem (dis)advantageous in the current adaptive landscape, but selection exhibits temporal variation

• Read here on why and how genetic variation is required for evolution/adaptation to occur

• Read here for more about the process of adaptation

• Read here for more on genetic variation

• Read here for more on rapid evolution

• Read here for more on the relationship between multivariate selection and adaptation

• A detrimental phenotypic character may have no genetic variance, so it may be impossible for selection to stop such occurring; many diseases have considerable environmental inducers (e.g. selection can't work against emphysema developed as a result of smoking, but it can work against genes that increase the tendency of developing emphysema).

Answer 6

All the previous answers are very good. However, I feel a point was missed (or maybe I didn't read deeply enough).

I will highlight the concept of fitness landscapes. This is how it looks:

The peaks represent the fitness of the species for a particular allele frequency. In the multipeak scenario, there are valleys which represent, obviously, reduced fitness. The concept of evolution being a chance event translates to "local peaks are favored over the global peak". As a result, a species may reach a high fitness level even though a higher peak exists (assume the highest peak is the perfect organism you speak of). Once it has climbed a peak, to change to the other peak it has to go through a valley i.e reduced fitness to improve. If nature was a benevolent entity then it could have been a grand transition. However, less fitness makes species susceptible to die off and hence they mostly don't make the transition making the perfect organism statistically less likely.
I also attach a link a .gif to show species heading to a certain peak. Notice the fact that no one likes to stay in the valley and rapidly swarms to the peak.

Answer 7

Richard Dawkins devoted an entire chapter of The Extended Phenotype to this question, Constraints on Perfection (the third chapter in the edition I have to hand); he listed six (not including those he critiqued as perhaps not being such strong constraints as others have suggested). The whole thing is worth a read, but for brevity, I'll state the six he had in mind here:

• Time lags (a given adaptation may arrive in the future)
• Historical constraints (the new system has to modify the old rather than starting from scratch)
• Available genetic variation (for example, this may explain why many non-functional body parts shrink instead of disappearing)
• Constraints of costs and materials (big brains, for example, are so expensive we're very unusual in having a net use for them)
• Imperfections at one level due to selection at another level (Dawkins had previously spent a whole book, The Selfish Gene, arguing - among other things - that what's good for the gene may not be good for the individual)
• Mistakes due to environmental unpredictability or 'malevolence' (an organism is built to be well-suited to average and dangerous conditions in their environment, and cannot quickly update its structure for every possible contingency)

Answer 8

Let's break this down to cover your two questions individually

Question 1 If a trait would be advantageous to an organism then why hasn't it evolved yet?

This one is really easy, natural selection, as well as other forms of selection, can only work with the variation that arises through mutation. Mutation is a random process evolution has no control about what variation will arise it can only select from those that do. Successful mutations tend to be small changes (small changes are less likely to disrupt a complex process like an organism) because of this organisms can get stuck with disadvantageous setups because evolution cannot go back to drawing board, it cannot go through a disadvantageous change to get an advantageous one.

My classic example is human back problems, a spine is a horrible thing to use to support weight upright but a spine was all evolution had to work with, it was the best option of what was available and being upright was a bigger advantage than the problems with a spine were a disadvantage. Designing a whole new system would be better but the chances of the hundreds of thousands of mutations need to "start over from scratch" are so astronomically unlikely we will never see it. Evolution can only select for the best out of what is available and mutation can only make small changes to what is available.

Another example is the Laryngeal nerve in giraffes which is a 14ft nerve that literally only goes a few inches, but since the nerve evolved when a straight line from the brain to the tongue left the heart in front of it (fish), as terrestrial vertebrates developed necks and moved the heart into the chest the nerve is stuck running down around the heart before returning to the head, a ~14ft detour. having the nerve is a bigger advantage than the detour is a disadvantage, and a mutation redirecting the nerve would be a large complex mutation thus is unlikely to ever occur. A historical legacy like this is responsible for a lot of bad adaptations that stick around, changing them is just too difficult to happen by random mutation.

Question 2 Conversely, if a trait is not advantageous or mildly disadvantageous, why does it exist?

there are a few components to this no trait is advantageous in an of itself, it is advantageous in a particular organism and environment. Sharks beautifully adapted masters of the ocean, drop one in the Sahara and it is doomed. what is good in one environment is often a disadvantage in another. Combine this with the fact that organisms move(or spread) and the fact that environments change and it is easy to see how organisms can end with a mildly disadvantageous adaptation. Adaptations for hunting from the ice was an advantage for polar bears when ice was plentiful now that the ice is disappearing it is not so advantageous. but the ice is disappearing faster than evolution can change the polar bears.

Often the changes in environment are changes in the other organisms around them, being extremely fast is actually a horrible adaptation for gazelle in many ways, it makes them fragile, it makes them small and weaker than other ungulates, it spends calories to grow muscles that could be used to produce larger offspring or taller bodies, etc, but in an environment with cheetah not being fast is a worse disadvantage. In this way, evolution is often about taking the least bad out of a set of bad "choices". And of course, the faster the gazelle get the faster the cheetah get until they are too specialized to live any other way. This push to specialization is very common however specialized organisms are the least able to handle change, they are too specialized to do anything different. So if say a new predator shows up that can force fragile cheetah away from their kills (humans) the cheetah are screwed, they could suffer a huge die off (Africa) or go extinct (North America). Evolution cannot make you good in every direction, you cannot be both a good swimmer and a good climber and a good digger.

Another aspect is cost vs benefit, a trait does not occur in isolation, a body is a complex thing, the same mutation that makes some humans resistant to cholera also makes them more susceptible to cystic fibrosis, this will be an net advantage when cholera is a big threat and a net disadvantage when cholera is rare. they key term here is net there is no condition when it is purely an advantage or disadvantage. Basically, all traits fall into this cost-benefit comparison. No trait is an advantage in every situation, every change has a cost, even if it is just calories that could be spent elsewhere. We know of no traits that are purely a disadvantage that persists but there are many many (if not all of them) such tradeoff traits, a disadvantage for an advantage.

Another example is sexual selection, some traits help you to find or attract mates but are detrimental to your survival, but reproduction is a bigger advantage in evolution than survival. Evolution simply cannot favour genes if they never make it into the next generation consistently. So you end with peacock tails that get the males killed but are the only way to find mates. A male without the huge tail will not mate and the male offspring of a female without the desire for big tails will suffer the same problem (because the female keeps ending up with unattractive male offspring) so it is basically impossible for peacocks males to stop growing huge tails and it is unlikely peacock females will stop preferring bigger and bigger tails. Thus even though it is very detrimental to peacock survival the trait sticks around because it is an advantage in reproduction.