Phenotypic variation in populations arises from both genetic differences and environmental variation. For selection to act, variation must be heritable—passed from parents to offspring. Not all variation is genetic; much may be environmental, making it invisible to natural selection. Understanding the genetic basis of variation is essential for predicting evolutionary responses.
From Mendelian genetics, you know that alleles segregate and assort independently, producing offspring with different genotypes. From the genetic code, you know that genotype encodes the proteins and regulatory signals that build an organism. But when you look at a real population — say, a field of wildflowers varying in height — you are seeing phenotypic variation, the combined product of genetic differences, environmental differences, and their interaction. Disentangling these sources is the central challenge that connects genetics to evolution.
Consider plant height. Some variation is clearly genetic: tall parents tend to produce tall offspring because they pass on alleles that promote growth. But some variation is environmental: a genetically identical clone planted in rich soil grows taller than one in poor soil. And some is genotype-by-environment interaction: a genotype that thrives in wet conditions may perform poorly in dry ones, while another genotype shows the opposite pattern. When you observe a population, all three sources are mixed together in the phenotypes you see. A plant that appears tall might carry "tall" alleles, or it might simply have landed in a favorable microhabitat.
This distinction matters enormously for evolution because natural selection can only act on heritable variation — the portion of phenotypic differences that is transmitted from parent to offspring through genes. If all the height variation in a population were caused by soil quality, selecting the tallest plants as parents would not produce taller offspring in the next generation, because the parents' advantage was environmental, not genetic. Conversely, if most variation is genetic, selection on the tallest plants will shift the population mean upward. The fraction of total phenotypic variation attributable to genetic differences is called heritability, a concept you will study in depth soon.
This is why population geneticists care so much about partitioning variance. Measuring phenotypic variation alone tells you nothing about evolutionary potential — you must know how much of that variation has a genetic basis. Breeding experiments, twin studies, and parent-offspring regressions all exist to answer this question. The practical implications are immediate: a crop breeder selecting for yield needs heritable variation to make progress; a conservation biologist predicting whether a species can adapt to climate change needs to know whether the relevant traits have genetic variation to work with. Without heritable phenotypic variation, natural selection has no raw material, and evolution stalls.