Heritability and fitness

By Razib Khan | October 11, 2010 3:09 pm

Words matter, and they can confuse. Here’s Wikipedia’s preamble for heritability:

Heritability is the proportion of phenotypic variation in a population that is attributable to genetic variation among individuals. Phenotypic variation among individuals may be due to genetic and/or environmental factors. Heritability analyses estimate the relative contributions of differences in genetic and non-genetic factors to the total phenotypic variance in a population.

Pay close attention to the variation part of “phenotypic variation”: if a trait has a heritability of 0.5, it means that the phenotypic variation is 50% due to genetic variation. It does not imply that the trait is 50% caused by genetics.

Heritability is also specific to a particular population in a particular environment.

Height and I.Q. are both heritable traits. Tendencies run in families. But the heritability is higher for both traits in developed societies or among those of middle class or higher. Why? Consider height. There are diminishing returns on nutritional input. Beyond a certain point eating more just makes you wider, not taller. In less developed societies where nutritional stress is more common there is a stronger environmentally induced socioeconomic correlation between height and status. European aristocrats who arrived in the United States in the 19th century were often shocked to meet American citizens of low or middling status who looked them eye-to-eye. Because of the surplus of land relative to labor in North America there was less nutritional stress than in Europe.

But the point of this post is to reiterate a rule-of-thumb which is useful to keep in mind: traits which are strongly correlated with fitness tend to be less heritable, in that less of the variance of the trait value within the population is correlated with genetic variance. Why? Because natural selection will eliminate that variation over time. After all it operates upon fitness differentials within populations across heritable traits. It tends toward equilibrium. Here’s a figure from the paper I pointed to earlier:


heritable

Age at first birth has a clear fitness implication. Fitness after all manifests through reproduction. Height on the other hand is a more confused issue. There may be antagonistic selection across the sexes, and, there is clearly a point where individuals who are too tall begin to suffer physiological problems. More later….

CATEGORIZED UNDER: Evolution, Genetics
MORE ABOUT: Evolution, Genetics
  • Josef Uyeda

    Compare this with David Houle’s (1992) comparing mean-standardized evolvability with heritability. Heritability suffers from the fact that it conflates two quantities. Therefore, a trait can decrease in heritability by two mechanisms. 1. It can decrease by decreasing the additive genetic variance, e.g. because it is fitness related and selection is reducing genetic variation. 2. It can decrease because of increasing phenotypic variation. This is what you’re talking about with the differences in heritability of height in different cultures. The additive genetic variation doesn’t change, the phenotypic variation does.

    However, Houle found that when standardized by the mean rather than the phenotypic variance, life history traits related to fitness actually had MORE genetic variance and were more evolvable than traits less tied to fitness (like height). That is, age at first birth may in fact be able to evolve faster than height. The reason that heritability is lower is because the phenotypic variance for “age at first birth” is much, much higher than that of height, because it is a more complex trait with a larger environmental component. However, more loci affect “age at first birth” and the trait actually may have more genetic variation and be more evolvable (I don’t know if that’s true for most traits, but the correlation between evolvability and heritability was basically 0). For those reasons, comparing heritability across different traits can actually lead to exactly the opposite conclusion than is warranted here.

    Perhaps this was the direction you were going with the “words matter” bit, and I just stole your thunder.

  • Chris T

    Very true, be suspicious of arguments using trait heritability estimates that do not specify which population it was measured in.

  • Pingback: Tweets that mention Heritability and fitness | Gene Expression | Discover Magazine -- Topsy.com()

  • Pingback: Heritability and fitness | Gene Expression | Discover Magazine | Health & Wellness()

  • Josh

    Why? Because natural selection will eliminate that variation over time. After all it operates upon fitness differentials within populations across heritable traits. It tends toward equilibrium. Here’s a figure from the paper I pointed to earlier:

    This isn’t necessarily true. A lot of traits that are highly correlated with fitness are “high level” traits, and I think that most quantitative geneticists would argue that there is simply more chance for environmental variation to get “in there”.

  • Pingback: Flashlinks | Evolving Thoughts()

  • http://blogs.discovermagazine.com/gnxp Razib Khan

    josh, i have some papers on the issue i’m looking over…it is more complicated than what i learned in the textbooks. would you recommend something specifically recent as representative? (obviously can’t read ALL the papers :-)

  • outeast

    Do transcription errors and the like count as ‘environmental variation’?

  • Divalent

    I’d like to extend the tangent “outeast” opened up about this idea that any non-genetic factor influencing a phenotype is due to “environment”. There really should be a third category to denote the variation due to random stochastic events that aren’t really due to “environment” as we commonly define that term. The basal error rates of enzymes, background mutations due to spontaneous radioactive decay, stochastic variations in the production of signaling chemicals that trigger abrupt start and stop points in the development process, etc.

    The problem with the term “environment” is that it often implies something that one can control or modify, or at least to characterize and predict. Although, like heritability under different circumstances, the proportion varies depending on the amount of genetic and true environmental influences, there will always be a component of what we currently call “environment” that can never be controlled or predicted.

  • Eric Johnson

    > the phenotypic variance for “age at first birth” is much, much higher than that of height

    Higher in what units?

  • Chris T

    I’d like to extend the tangent “outeast” opened up about this idea that any non-genetic factor influencing a phenotype is due to “environment”. There really should be a third category to denote the variation due to random stochastic events that aren’t really due to “environment” as we commonly define that term. The basal error rates of enzymes, background mutations due to spontaneous radioactive decay, stochastic variations in the production of signaling chemicals that trigger abrupt start and stop points in the development process, etc.

    This is something that annoys me too. Too often heritability estimates are given and then it’s said that it leaves the rest to be explained by environmental influences often implying that we can control it. Well no, genes do not get perfectly expressed and early random alterations in development can lead to big down downstream changes. That’s a major reason why ‘identical’ twins are not exactly alike.

  • Josh

    Razib,

    Hrmm…. it’s not really my area of expertise… This seems to be one of the standard references, but I haven’t read it.

    http://www.genetics.org/cgi/content/abstract/130/1/195

  • nick1

    re fitness traits: http://www.nature.com/hdy/journal/v83/n2/pdf/6885850a.pdf

    “Fitness traits seem to have higher levels of additive genetic variance than nonfitness traits – an observation that has been explained in terms of the larger number loci influencing fitness as compared to nonfitness traits.”

  • Josef Uyeda

    Eric Johnson Says:
    the phenotypic variance for “age at first birth” is much, much higher than that of height
    Higher in what units?

    Right, I was careless when I wrote that and should have said “could be” rather than “is”, since I’m not actually working with data. Comparing measurements for different classes of traits is often very difficult, and sometimes impossible, but I was going for something like the Coefficient of Variation. It’s very easy to run into the classic problem of comparing apples and oranges (only worse, apples and oranges are actually pretty easy to compare).

    Regarding what Chris T says about heritability. We’re talking about narrow-sense heritability, which only counts additive genetic variance. Nobody is arguing that the environmental component is entirely “external environment” and doesn’t involve anything genetic (compare to broad-sense heritability). The implication isn’t that the environmental component is controllable and predictable, it’s that it’s unpredictable (or possibly too complex or inconsequential to model).

  • Chris T

    Nobody is arguing that the environmental component is entirely “external environment” and doesn’t involve anything genetic (compare to broad-sense heritability). The implication isn’t that the environmental component is controllable and predictable, it’s that it’s unpredictable (or possibly too complex or inconsequential to model).

    When conveying the subject to a lay audience, these qualifiers are generally left off and people generally interpret it as meaning ‘controllable’. People working in the field may grasp the nuance; most non-professionals will not.

NEW ON DISCOVER
OPEN
CITIZEN SCIENCE
ADVERTISEMENT

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at http://www.razib.com

ADVERTISEMENT

See More

ADVERTISEMENT

RSS Razib’s Pinboard

Edifying books

Collapse bottom bar
+

Login to your Account

X
E-mail address:
Password:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »