Recently I worked quite heavily on figuring out how to extract variance components for analyzing Qst. Here is a summary of the useful web resources that I found helpful. The between-population and within-population variance can be extracted from phenotypic data mainly in 3 ways.
The most straightforward approach is to use ANOVA. For how to use ANOVA to partition variance at between- and within-population level, I suggest to read page 556-559 (especially Table 18.1) in Lynch and Walsh (1998) – Genetics and analysis of quantitative traits.
It is also possible to use liner mixed models to partition variance. We can use the lmer package or use the MCMCGlmm package. I came across several useful webpages for introducing how to set up the linear model in these packages and how to interpret the analysis results.
First, a good introduction to linear mixed model and fixed and random effects can be found here. A simple example on how to extract variance at different levels can be found here. In case if you are confused about variance components in linear mixed models, this webpage provides a good explanation of variance in linear mixed models using simulation approach. For how to extract variance components from lmer results, this is the best explanation I came across.
Regarding MCMCGlmm package, which provides a MCMC approach for fitting a generalized linear mixed model to your data, this tutorial provides a good resource for how to set up a model and specify priors. If you want to see a comprehensive guide for prior setting in MCMCGlmm, please check out the course notes for MCMCGlmm.