Write notes on/Answer the following in about 150 words each :
Selection for threshold traits is often frequency-dependent.
10 marks · answer in 150 words · directive: Write notes
The core of a good answer
For all-or-none traits controlled by an underlying continuous liability and a threshold, the observed heritability and the change in incidence per generation depend on the incidence itself: response is greatest near 50% incidence and very slow when the trait is rare or almost universal.
An outline written by upsc.help, not an official UPSC answer; every fact checked by a separate review. How we work
What the question asks
Explain the liability threshold model and show why the effectiveness of selection against (or for) a threshold character depends on how frequent the character is in the population. A note of about 150 words.
Answer outline
Introduction
Threshold traits (twinning, calving difficulty, mastitis, many congenital defects, survival) appear in two or a few classes but are polygenic; Wright and later Falconer explained them by a normally distributed liability with a threshold above which the trait is expressed.
The model
- Liability = genetic + environmental effects, normally distributed; animals beyond the threshold show the trait.
- Incidence p is the area of the curve beyond the threshold; heritability is estimated on the liability scale.
Why selection depends on frequency
- A shift in mean liability of Δ (in SD units) changes incidence by about z·Δ, where z is the height of the normal curve at the threshold.
- z = 0.399 at 50% incidence but only 0.103 at 5% and 0.027 at 1%: the same genetic gain gives a much smaller visible change in a rare trait.
- Heritability on the observed 0/1 scale = h² (liability) × z²/[p(1 − p)] (Dempster and Lerner): the factor is 0.64 at 50%, 0.22 at 5% and 0.07 at 1% incidence.
- Selection differential on liability also depends on incidence: when few animals are affected, culling them removes little of the liability variation.
Practical implications
- Selection against rare defects or for rare twinning by phenotype alone is slow; use family and progeny records or genomic predictions on the liability scale.
Conclusion
Because response measured in incidence depends on where the threshold lies relative to the population mean, selection for threshold traits is frequency-dependent, and breeders should work on the underlying liability rather than the observed classes.
Examples and data to cite
- Falconer introduced the liability threshold model in his textbook (1960) and papers in 1965 and 1967; Sewall Wright earlier (1934) used a threshold to explain the extra hind toe in guinea pig strains.
- Dempster and Lerner (1950) gave the conversion of heritability between the observed binary scale and the underlying liability scale.
Facts to remember
- Change in incidence ≈ z × change in mean liability (z = normal ordinate at the threshold).
- z values: 0.399 (p = 0.5), 0.280 (p = 0.2), 0.103 (p = 0.05), 0.027 (p = 0.01).
- h²(observed) = h²(liability) × z²/[p(1 − p)].
- Threshold traits are polygenic, not single-gene, despite discrete expression.