The dihybrid cross

Concept

The dihybrid cross

T-014Home BU-103Threads information · evolution

Statement

Predicting the 9:3:3:1 ratio of two traits.

Why it matters

law-of-segregation established that a single gene's two alleles separate cleanly into different gametes; the dihybrid cross is the direct test of what happens when two such genes are tracked simultaneously, and its characteristic 9:3:3:1 ratio is the empirical signature of independent-assortment, the law it experimentally establishes. Understanding why this specific ratio arises, purely from probability applied to two independently segregating gene pairs, is also the necessary foundation for later recognising when genes deviate from it — the departures linkage-recombination and sex-linkage go on to explain.

Hypotheses

The two genes under study assort independently of one another, i.e. lie on different chromosomes (or are far enough apart on the same chromosome to segregate as if unlinked).Independent assortment is precisely the claim the dihybrid cross ratio is used to test; genes physically close together on the same chromosome tend to be inherited together rather than independently, producing a different, non-9:3:3:1 ratio (linkage-recombination), so this assumption is not automatic and must hold for the standard ratio to be expected. Each gene shows simple, complete dominance, with one allele's phenotype fully masking the other's in the heterozygote.The characteristic 9:3:3:1 phenotypic ratio specifically assumes only four distinguishable phenotypic classes result from two genes, each with two alleles; incomplete dominance, codominance, or epistasis between the two genes each alter the observed phenotypic ratio even while the underlying genotypic ratio remains unchanged. Sample size must be large enough for the ratio to be statistically distinguishable from alternative hypotheses (such as linkage); small crosses can show considerable random deviation from the expected 9:3:3:1 ratio purely by chance, which is why a chi-squared test is the standard tool for evaluating whether an observed ratio is consistent with independent assortment.

Proof

\text{A dihybrid parent } (AaBb) \text{ produces four gamete types } (AB, Ab, aB, ab) \text{ in equal proportion, } \tfrac14 \text{ each.}

This follows directly from applying the law of segregation (law-of-segregation) independently to each of the two gene pairs, then combining the two independent 1:1 segregation outcomes; because the two genes assort independently (Hypotheses), every combination of one allele from each gene is equally likely. A

P(\text{offspring genotype}) = P(\text{allele from parent 1}) \times P(\text{allele from parent 2})

Because fertilisation combines gametes at random, and the two genes' allele identities are themselves independent (Step 1), the joint probability of any specific two-gene offspring genotype is simply the product of the two genes' individual segregation probabilities — the same multiplication rule of independent probability used throughout Mendelian genetics. A

\text{A 16-cell Punnett square, combining all 4} \times \text{4 gamete pairings, gives the full genotypic distribution of an } AaBb \times AaBb \text{ cross.}

Systematically pairing each of the four gamete types from one parent with each of the four from the other (Step 1) generates all 16 equally likely genotype combinations; grouping these by resulting phenotype, given complete dominance at each gene (Hypotheses), yields exactly four phenotypic classes. A

9\,A\_B\_ : 3\,A\_bb : 3\,aaB\_ : 1\,aabb

Counting the 16 genotype combinations of Step 3 by phenotype: 9 combinations show both dominant traits, 3 show the first dominant with the second recessive, 3 show the reverse, and exactly 1 (the doubly homozygous recessive) shows both recessive traits — the characteristic 9:3:3:1 ratio. A

Result

AaBb \times AaBb \ \longrightarrow\ 9\,A\_B\_ : 3\,A\_bb : 3\,aaB\_ : 1\,aabb

Reading. When two genes assort independently and each shows complete dominance, crossing two double heterozygotes produces exactly four phenotypic classes in the fixed ratio 9:3:3:1, a direct consequence of combining two independent 3:1 monohybrid ratios.

Scope. Requires independent assortment (unlinked genes) and complete dominance at both loci (Hypotheses); linked genes, or genes showing epistasis or incomplete/codominance, produce systematically different ratios, discussed further below.

Corollaries & converses

The 9:3:3:1 ratio can equivalently be derived as the product of two independent 3:1 monohybrid ratios, \((3:1)\times(3:1)\), directly reflecting Step 2's multiplication rule and confirming that a dihybrid cross adds no new genetic principle beyond independent-assortment applied twice.
Epistasis (one gene's phenotype masking or modifying another's) modifies the 9:3:3:1 ratio into other characteristic fixed ratios (commonly 9:3:4, 12:3:1, 9:7, and others), each pattern diagnostic of a specific type of gene interaction, while still resting on the same underlying 9:3:3:1 genotypic-class proportions.
Converse: an observed phenotypic ratio that deviates substantially and consistently from 9:3:3:1 in a dihybrid cross is evidence against independent assortment for that gene pair (i.e. evidence of linkage, linkage-recombination) or evidence of a gene interaction (epistasis) altering the expected phenotypic classes, and a chi-squared test is the standard means of distinguishing genuine deviation from ordinary sampling variation.

Fails without

Drop independent assortment (Hypotheses), i.e. the two genes are linked: gametes are no longer produced in the equal 1:1:1:1 ratio of Step 1; parental-type gamete combinations become overrepresented and recombinant types underrepresented, skewing the resulting phenotypic ratio away from 9:3:3:1 toward an excess of the two parental phenotype classes, exactly the signature linkage-recombination uses to detect and measure linkage.
Drop complete dominance (Hypotheses), i.e. one or both genes show incomplete dominance or codominance: heterozygotes at that gene no longer resemble the homozygous dominant phenotype, so more than four phenotypic classes result even though the underlying 9:3:3:1 genotypic ratio is completely unchanged — the ratio's numeric pattern is a statement about genotype classes filtered through complete dominance, not an intrinsic property of the genotypes alone.

Common errors

Assuming every dihybrid cross must produce a 9:3:3:1 phenotypic ratio regardless of dominance relationships or linkage; the ratio specifically requires both independent assortment and complete dominance at each gene (Hypotheses).
Constructing an incorrect Punnett square by combining alleles gene-by-gene rather than gamete-by-gamete, producing an incomplete or incorrectly weighted set of genotype combinations rather than the full, correctly enumerated 16-cell grid of Step 3.
Interpreting any deviation from 9:3:3:1, however small, as proof of linkage without first applying a chi-squared test to check whether the deviation exceeds what ordinary sampling variation would produce (Hypotheses, t3).
Confusing the genotypic ratio (9:3:3:1 for the specific class groupings shown in Step 4, though there are in fact 9 distinct genotypes overall) with the phenotypic ratio; the two coincide numerically here only because of the specific grouping complete dominance produces.

Discussion

Gregor Mendel's own pea-plant experiments, published in 1865, included dihybrid crosses (tracking seed shape and seed colour together) specifically to test whether the segregation of one trait pair was independent of another; the resulting 9:3:3:1 ratio he observed became one of the two foundational laws of classical genetics, later formalised as the law of independent assortment (independent-assortment). Mendel's success is now understood to have partly depended on his choice of trait pairs, several of which happen to lie on different pea chromosomes or far enough apart on the same chromosome to assort essentially independently, sparing him from complications linkage would otherwise have introduced.

Because linkage was unknown in Mendel's time, his choice of seven trait pairs across a genome of seven chromosome pairs, none showing detectable linkage, has sometimes been noted as a fortunate (or, on some readings of the pea genome, partly coincidental) circumstance that allowed a clean, generalisable independent-assortment result to emerge from his data.

Common misconception: that the number "16" in a dihybrid Punnett square and the numbers "9, 3, 3, 1" refer to the same thing. The 16 cells represent equally likely individual genotype combinations (Step 3); the 9:3:3:1 ratio is these 16 outcomes regrouped into four phenotypic classes under complete dominance (Step 4) — the genotypic diversity underlying the cross is considerably richer (9 distinct genotypes) than the four phenotypes that ratio describes.

Worked examples

\text{Pea plants: } R\ (\text{round seed, dominant}),\, r\ (\text{wrinkled}); \ Y\ (\text{yellow, dominant}),\, y\ (\text{green}). \ \ RrYy \times RrYy

Each gene independently follows a 3:1 monohybrid ratio (\(3\,\text{round}:1\,\text{wrinkled}\); \(3\,\text{yellow}:1\,\text{green}\)); since the two genes assort independently (different chromosomes in pea), the combined phenotypic ratio is the product of the two monohybrid ratios. A

(3:1)\times(3:1) = 9\ \text{round yellow} : 3\ \text{round green} : 3\ \text{wrinkled yellow} : 1\ \text{wrinkled green}

Reading. The classic 9:3:3:1 ratio, applied to Mendel's own round/wrinkled and yellow/green pea traits, reproduces exactly the historical result that established independent assortment as a general genetic principle.

Scope. The identical calculation applies to any pair of independently assorting genes, each showing complete dominance, regardless of the specific organism or trait pair involved.

Problems

A dihybrid cross between two double heterozygotes for unlinked genes produces 320 offspring. Using the 9:3:3:1 ratio, calculate the expected number of offspring in each of the four phenotypic classes.
Solution
Total ratio parts \(=9+3+3+1=16\); each part corresponds to \(320/16=20\) offspring. Expected numbers: \(9\times20=180\) (both dominant), \(3\times20=60\) (first dominant, second recessive), \(3\times20=60\) (first recessive, second dominant), \(1\times20=20\) (both recessive).
A dihybrid cross produces offspring in a ratio much closer to 1:1:1:1 among the four expected phenotypic classes than to 9:3:3:1. Propose an explanation using the Hypotheses, and state what additional statistical test would help confirm it.
Solution
A ratio distorted toward 1:1:1:1 rather than 9:3:3:1 is not the classic signature of linkage (which instead skews toward excess parental types, typically giving ratios closer to the two extreme classes than the standard 9:3:3:1 predicts); more directly, this pattern is inconsistent with the assumption of complete dominance and independent assortment together producing four unequal classes, so it is worth checking whether the cross is in fact a testcross (heterozygote \(\times\) double homozygous recessive) rather than the dihybrid \(\times\) dihybrid cross of Step 3–4, since a testcross of unlinked genes is expected to give a 1:1:1:1 ratio directly. A chi-squared test against each candidate expected ratio (9:3:3:1 vs 1:1:1:1) would confirm which cross type the data actually fit.
Explain why observing a dihybrid cross's phenotypic ratio alone cannot distinguish between "the two genes are linked with 50% recombination frequency" and "the two genes are on different chromosomes," even though both give a 9:3:3:1 ratio.
Solution
A recombination frequency of 50% means recombinant and parental gamete types are produced in exactly equal proportion, which is statistically indistinguishable from true independent assortment (Step 1's equal 1:1:1:1 gamete ratio) — both produce the identical 9:3:3:1 phenotypic ratio in the following generation. Distinguishing true independent assortment (genes on different chromosomes) from linkage with 50% recombination (genes on the same chromosome but far enough apart to assort as if unlinked) requires additional evidence beyond the dihybrid cross ratio itself, such as direct chromosomal mapping.