SYNOPSIS
Statistical testing tools are fundamental for data analysis in plant breeding and helps in selection and improvement of genotypes. They aid breeders to efficiently assess variability across environments and seasons, which ensures to draw reliable inferences.
Key statistical tools include measures of central tendency (mean, median and mode) to summarize trait performance and measures of dispersion (range, variance, standard deviation and coefficient of variation) to assess variability and experimental precision. Skewness and kurtosis also known as third and fourth degree statistics respectively helps to interpret shape of data distribution. First and second degree statistics aids in analysis of quantitative traits.³
The study in an F₂ population of tomato (‘Arka Vikas’ × ‘Red Ball’) revealed significant genetic variability, with key traits like fruit firmness and pericarp thickness positively influencing shelf life. Skewness and kurtosis analyses indicated the presence of dominance and epistatic gene action, suggesting that both additive and dominance effects can be effectively utilized for improving shelf life in tomato breeding programs.²
Probability distribution plays a key role in modelling and analyzing trait performance in plant breeding. Quantitative traits often follows normal distribution, which is a pre-requisite for many statistical techniques. Binary traits such as disease resistance are modeled using the binomial distribution, while count data like the number of pests or tillers fit the Poisson distribution. Depending on the data type and distribution, breeders choose between parametric and non-parametric tests. Parametric tests like the t-test, F-test, Z-test, ANOVA and regression methods are preferred for normally distributed and continuous data. When these assumptions are violated, non-parametric tests such as chi-square (χ²) tests, Mann-Whitney U tests, Wilcoxon signed-rank or Kruskal-Wallis tests are preferred due to their robustness to outliers and non-normality.³
Genetic analysis of growth habit in dolichos bean has been revealed 15:1 F₂ segregation ratio in photoperiod-insensitive (PIS) backgrounds and a 9:7 ratio in photoperiod-sensitive (PS) backgrounds. These ratios were confirmed in F₃ generations through χ² test, indicating duplicate gene action in PIS and complementary gene action in PS backgrounds suggesting that the genetic control of growth habit varies with the degree of photoperiod sensitivity.¹
Crucial aspect of statistical analysis is hypothesis testing, where a null hypothesis (H₀) assumes no significant difference between genotypes and an alternative hypothesis (H₁) suggests a difference. Based on p-values and a significance level (typically α = 0.05), breeders accept or reject H₀ to make objective, data-driven decisions. Therefore, statistical parameters were vital for estimating variability, validating results and guiding the selection of superior genotypes in plant breeding.
REFERENCES:
¹ BASANAGOUDA, G., RAMESH, S., CHANDANA, B.R., SIDDU, C.B., KIRANKUMAR, R. AND KALPANA, M.P., 2022. Inheritance of growth habit under photoperiod insensitive genetic background in dolichos bean (Lablab purpureus (L.). Genet Resour Crop Evol., 69 (7): 2535-2543
² PAVAN, M.P., GANGAPRASAD, S. AND ADIVAPPAR, N., 2022. Genetic variability, trait inter-relationships, third and fourth degree statistics based genetics for fruit quality and yield traits governing shelf life in tomato (Solanum lycopersicum L.). Plant Genet. Res., 25(05): 366-375.
³ RANGASWAMY R., 2024. A textbook of Agricultural Statistics 4th edition : 01-791
%20(2).png)
0 Comments