- Background -

Talking about one phenotypic trait, the total phenotypic variance $V_p$ is decomposed into genetic and environmental variance for this trait, represented by the symbols $V_G$ and $V_E$ respectively. Also the covariance between genes and environment affect $V_p$ but we'll ignore it for the purpose of this question.

$$V_p = V_E + V_G + O(Cov(E,G))$$

The genetic variance $V_G$ can be further decomposed into dominance and additive variance. Similarly the environmental variance can be further decomposed into for example, for the mammals, the variance in the womb $V_W$ and the other environmental variance $V_{OE}$.

$$V_p = V_w + V_{OE} + V_A + V_D$$

$$h_N^2 = \frac{V_A}{V_p} = \frac{V_A}{V_w + V_{OE}}$$

One could measure $V_W$ by looking at correlation between dizygotic twins.

- Question -

What part of the environmental variance in phenotype is explained by the variance in wombs' environments? i.e. what are the estimates of $\frac{V_w}{V_e}$ or what are the estimates of $\frac{V_w}{V_p} = \frac{V_w}{V_w + V_{OE}}$?

Of course the answer will depends on the trait and on the species. I welcome either a general idea/estimation or any such measure for any specific trait.

  • $\begingroup$ Variance in womb is dependent on environmental variance. Some environmental effects are suppressed (some inflammatory modulators) whereas others are not. I am afraid your question is too broad. Moreover you are bringing in a lots of concepts together. $\endgroup$
    Commented Dec 24, 2014 at 12:35
  • $\begingroup$ The womb environment depends on the non-womb environment (and probably on the genetics). Yes, there are covariance terms between $V_w$ and $V_E$ (and probably between $V_w$ and $V_G$) just like there are covariance terms between $V_E$ and $V_G$. It doesn't change the fact that there is a true $V_w$. My question might eventually be too broad because I don't specify a given phenotypic trait of interest. I edited my question to welcome answers about any phenotypic trait. I don't fully understand why my question mix up many different concepts? It sounds pretty straight forward to me. Thks $\endgroup$
    – Remi.b
    Commented Dec 25, 2014 at 10:34
  • $\begingroup$ By "a lot of concepts" I mean that this question can be asked in a rather simple way. It would be straightforward to you because you posed it and you work on the area of population-biology. If I am not wrong this question has more to do with physiology (basically how well is a mammalian embryo protected against the environmental changes). Now it depends what those environmental changes are- temperature, chemicals (drugs), inflammation etc. $\endgroup$
    Commented Dec 25, 2014 at 10:43
  • $\begingroup$ Now the second part is whether these changes that are relayed (with some attenuation?) to the embryo, will have an effect on some trait. So there are several environmental effects. You can break the question down that way; nonetheless you have to specify some environment variable else it gets too broad. Trait is still predictable if you fix the environmental factor. $\endgroup$
    Commented Dec 25, 2014 at 10:48
  • $\begingroup$ Standard studies that aim at measuring the underlying genetic/environmental variances of variance in phenotypic trait do not look at the physiological mechanisms but only tries to determine the statistical associations. Also, typically these studies measure heritability through parent-offspring regression but does not make any consideration of specific environmental variables. You talk about "the second part of my question". Are there two parts in my question? $\endgroup$
    – Remi.b
    Commented Dec 25, 2014 at 10:53


You must log in to answer this question.