How can repeated measurements be used to distinguish between temporary and permanent environmental effects in livestock?
15 marks · directive: Explain
The core of a good answer
Repeated records on the same animal (successive lactations, fleece weights, litter sizes) let the phenotypic variance be split into variance between animals (genetic plus permanent environment) and variance within animals (temporary environment); their ratio, repeatability r = (VG + VEp)/VP, measures the permanent part.
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 variance partition behind repeatability, how it is estimated from repeated records, and how it is used to predict future performance and to improve selection accuracy.
Answer outline
Introduction
Many traits are expressed several times in an animal's life; differences that persist across records come from genes and permanent environment, while differences that vary from record to record come from temporary (special) environment.
Model and variance partition
- Record = μ + G + Ep + Et, where G is genotype, Ep the permanent environmental effect and Et the temporary effect of a single record.
- VP = VG + VEp + VEt; VEg (general) = VEp is shared by all records of an animal, VEs (special) = VEt differs between records.
- Repeatability r = (VG + VEp)/VP; it generally sets the upper limit of heritability.
Estimation
- One-way analysis of variance of records grouped by animal; r is the intra-class correlation σ²between/(σ²between + σ²within).
- Within-animal mean square estimates VEt; the between-animal component estimates VG + VEp.
- Mixed models (animal model with a permanent environment effect) separate VG from VEp when relatives are recorded.
Examples of the two kinds of effect
- Permanent: early life nutrition, udder damage after mastitis, loss of a teat, stunting by disease in calfhood.
- Temporary: season and year of calving, feed availability in one lactation, a short illness, management changes.
Uses
- Most Probable Producing Ability: MPPA = H̄ + [nr/(1 + (n − 1)r)](P̄ − H̄), for culling cows on their own records.
- Variance of the mean of n records relative to one record = [1 + (n − 1)r]/n: with r = 0.5, two records give 0.75 and four give 0.625, so averaging helps most when r is low.
- High r means one record predicts later ones well; low r means several records are needed before culling or selection.
Conclusion
By comparing variation within animals with variation between animals, repeated records show how much of performance is a lasting property of the animal, which guides culling, early selection and the number of records worth collecting.
Examples and data to cite
- Repeatability is the proportion of phenotypic variance due to differences between repeated measures of the same individual; it can be estimated only for traits expressed several times in life, and generally, but not always, it marks the upper level of heritability.
Facts to remember
- r = (VG + VEp)/VP ≥ h².
- Accuracy gain from n records: √(n/[1 + (n − 1)r]) relative to one record (1.15 for n = 2, 1.26 for n = 4 when r = 0.5).
- MPPA regression factor = nr/[1 + (n − 1)r].
- Temporary environmental variance is averaged out by multiple records; permanent variance is not.