Traditional ANOVA assumes all effects (except error) are fixed. However, in plant breeding, many effects (e.g., genotypes in a germplasm collection) are random—they are a sample from a larger population. Mixed linear models handle both fixed (e.g., environments, blocks) and random (e.g., genotypes, genotype × environment interaction) effects.
BLUP (Best Linear Unbiased Prediction) is the gold standard for predicting genetic merit. BLUP shrinks extreme estimates toward the population mean, accounting for differing numbers of observations and relationships. It is superior to BLUE (Best Linear Unbiased Estimation) when data are unbalanced or when genotypes are related. BLUP is integral to genomic selection (GS), where thousands of markers are used to predict breeding values.
Modern plant breeding has evolved from an art of visual selection into a rigorous science driven by quantitative genetics and applied statistics. The seminal works of researchers like Jawahar R. Sharma emphasize that without biometrics—the application of statistical methods to biological phenomena—the genetic improvement of crops would be slow, inefficient, and largely unpredictable. This essay outlines the key statistical and biometrical techniques fundamental to plant breeding, as reflected in comprehensive texts on the subject.
Biometrical genetics quantifies how much of the observed variation is heritable. Two forms are critical:
Genetic advance (GA) predicts the improvement expected from selecting a certain proportion of the population. The formula (GA = k \cdot h^2_n \cdot \sigma_P) (where (k) is selection intensity and (\sigma_P) is phenotypic standard deviation) guides breeders in choosing which traits and which selection intensities will yield progress.
This is where Sharma truly shines. While correlation tells you that yield and plant height move together, Path Analysis tells you why. Traditional ANOVA assumes all effects (except error) are
| Parameter | Formula | Significance | | :--- | :--- | :--- | | CV (Coefficient of Variation) | $(\sigma / \barx) \times 100$ | Measures precision of the experiment. | | Heritability (Narrow Sense) | $V_A / V_P$ | Reliability of selection. | | Genetic Advance | $K \cdot \sigma_p \cdot h^2$ | Actual gain expected. | | GCA Effect | $\textGeneral Mean - \textParent Mean$ | Additive gene action (breeding value). | | SCA Effect | $\textHybrid Mean - \textExpected Mean based on GCA$ | Non-additive gene action (hybrid vigor). |
This structured overview covers the primary "solid content" typically found in the chapters of Dr. Jawahar R. Sharma's book. It serves as a comprehensive guide for students preparing for exams (like ARS/NET) or breeders designing experiments.
Jawahar R. Sharma’s "Statistical and Biometrical Techniques in Plant Breeding"
is widely considered a cornerstone text for students and researchers in agricultural sciences. It bridges the gap between complex mathematical theory and the practical needs of a plant breeder.
Here is a breakdown of why this work remains a vital resource: 1. The Core Objective The book focuses on quantitative genetics Genetic advance (GA) predicts the improvement expected from
, providing the statistical tools necessary to understand how traits are inherited and how they can be improved. It moves beyond simple Mendelian genetics into the "messy" world of continuous variation—where traits like yield, height, and drought resistance are controlled by multiple genes and influenced by the environment. 2. Key Techniques Covered
Sharma meticulously details several essential biometrical methods, including: Analysis of Variance (ANOVA):
The foundation for partitioning phenotypic variation into genetic and environmental components. Mating Designs:
In-depth looks at Diallel, Line x Tester, and North Carolina designs to estimate combining ability and gene action. Stability Analysis:
Tools like the Eberhart and Russell model to see how varieties perform across different locations and years. Multivariate Analysis: Line x Tester
Using D² statistics and cluster analysis to measure genetic divergence, helping breeders pick diverse parents for hybridization. 3. Practical Utility What sets Sharma’s approach apart is the step-by-step application
. Instead of just presenting formulas, the text often guides the reader through data sets, showing how to interpret results to make actual breeding decisions (e.g., "Should I use mass selection or pedigree selection for this specific population?"). 4. Why it Matters Today
I cannot produce or provide a PDF copy of the book Statistical and Biometrical Techniques in Plant Breeding by Jawahar R. Sharma due to copyright restrictions. However, I can offer a detailed, original essay summarizing the core statistical and biometrical methods covered in such a text, which should serve as a study or reference guide.
Below is an essay on the subject, structured like an academic overview.