Meiosis, variation and reading a pedigree

Three separate sources of variation

Meiosis produces variation in three ways, and a good answer names which one it is talking about. Crossing over swaps segments between homologous chromosomes in prophase I, so a chromosome that came from one parent leaves carrying alleles from the other. Independent assortment lines up each homologous pair without regard to the others in metaphase I, so which parental chromosome goes to which pole is decided separately for every pair. Random fertilisation then combines one gamete out of an enormous number with another gamete out of an enormous number.

Only the first two happen during meiosis. The third happens after it. Questions asking for the variation produced by meiosis are asking for the first two; questions about variation in offspring generally want all three, and the marks are usually allocated one per mechanism.

Mitosis produces none of this, and the reason is worth stating rather than assuming: there is no pairing of homologous chromosomes, so there is nothing to cross over and nothing to assort. Each daughter cell gets a copy of every chromosome the parent cell had.

The comparison question, answered properly

Compare meiosis and mitosis and the marks are not for a table of differences. They are for differences that matter to the outcome, tied to the purpose of each process.

Mitosis makes two cells genetically identical to the parent, which is what growth and repair require - you do not want a liver cell improvising. Meiosis makes four cells with half the chromosome number and no two alike, which is what sexual reproduction requires, because two gametes fusing must not double the chromosome number every generation.

Say the chromosome number halves and say why that matters. An answer that lists two divisions, four daughter cells and crossing over, without ever connecting them to why the organism needs each process, is a correct answer that has not answered the question.

Setting out a cross without guessing

Write the genotypes before you draw anything. Decide what the letters mean and say so - a single line defining B as the allele for black fur and b for brown costs nothing and is often worth a mark on its own.

Then the Punnett square, then the ratio. Keep genotype ratios and phenotype ratios apart in your head and on the page. A monohybrid cross between two heterozygotes gives genotypes in a one to two to one ratio and phenotypes in a three to one ratio, and the question will specify which it wants.

A test cross is the one that catches people. You have an individual showing the dominant phenotype and you do not know whether it is homozygous or heterozygous, so you cross it with a homozygous recessive. If any offspring shows the recessive phenotype, the unknown parent carried the recessive allele. If none does, it probably did not - and probably is the honest word, because a small number of offspring proves very little.

• Define your symbols before you use them.
• Say whether a ratio is genotypic or phenotypic.
• Expected ratios are expectations, not promises; real litters are small.
• For sex linkage, write the alleles on the X, not beside it - X with a superscript, and a bare Y.

A method for pedigrees that works every time

Pedigree questions look like puzzles and they are not. There is an order of operations, and if you follow it you will get the answer without staring.

First, look for an affected child with two unaffected parents. That combination proves the condition is recessive, because the parents must both have carried it while showing nothing. If instead every affected individual has an affected parent, dominant is the likely answer, though it is rarely proved outright.

Second, once you know recessive, check sex linkage. An affected daughter with an unaffected father rules out X-linked recessive, because she would have had to inherit a recessive allele on her father's only X, and he would therefore have been affected himself. Affected individuals being overwhelmingly male points towards X-linked; a roughly even split points towards autosomal.

Third, only now, fill in genotypes. Start with the individuals whose genotype is forced - anyone affected by a recessive condition is homozygous recessive, and every parent of such a person carries at least one copy. Work outwards from those. Anyone whose genotype is still not forced gets written with a dash, and saying so is a better answer than picking one.

Where it usually goes wrong

The knowledge is rarely the problem in this topic. The habits are.

• Using genotype and phenotype as if they were interchangeable. They are the two things the question is distinguishing between.
• Assuming a condition is dominant because it appears in every generation of a small family tree.
• Writing a Punnett square before deciding what the letters mean.
• Explaining that meiosis produces variation without naming which mechanism - the marks are attached to the mechanisms.
• Forgetting that in an X-linked condition a male has only one allele, so there is no carrier state for him.

What to practise next

Do five pedigrees in a row using the order above - recessive or dominant first, sex linkage second, genotypes last - and do not let yourself skip to the genotypes. The order is the skill; the answers are a by-product.

Then write one paragraph, from memory, explaining why meiosis produces variation and mitosis does not. If you can do that in your own words without a diagram, you have the part of this topic the long-response questions are actually testing.

Check yourself

Three questions on what is above. Have a go before you open them - reading an answer you have not tried to give is the version of this that does nothing.