Agricultural Research Management Information System

Home
Research Summary
Search
About Us
- ARMIS
- Brochure
Contact Us
Report
User Request
Data Input
Help
- Operation Manual
  - PDF
  - Video
- Program Area & Commodity
We have reached 37600 number of research entries at this moment.

- Logout

Research Detail

Home
Research
Detail

Multivariate Analysis of BRRI Varieties in 2001-02

Dr. Debi Narayan Rudra Paul
PSO and Head
Agricultural Statistics Division

Abstract/ Executive Summary:

There exists strong ‘GxE” interaction for BRRI varieties --.the varieties are more responsive to environment in the Boro than in the T. Aman season.Mahalonobis D²is better than canonical roots method for classification of genotypes and distinguished the varieties exclusively based on only six characters: days to 50% flowering, 1000 grain weight, plant height, panicle length, Spikelete/panicle and flag leaf area at dough stage.Among six characters: days to 50% flowering, 1000 grain weight, plant height, panicle length, Spikelete/panicle and flag leaf area at dough stage, the contribution of days to 50% flowering to the divergence of genotypes is the highest.

Keywords: BRRI varieties, GxE interaction and Multivariate

Location: BRRI HQ, Gazipur

Start Date:

End Date:

Program Area: Socio-economic and Policy

Commodity: Rice

Objectives:

determine the relative contribution of phenotypic characters to rice yield,
classify BRRI varieties based on phenotypic characteristics and
identify the factors causing similarity and diversity among the genotypes.

Methodology:

The experimental design used was RCB with three replications. The number of varieties used was 15 in the T Aman and 20 in the Boro season Each plot size was 6- x 4-m² and spacing was 20- x 20-cm. Recommended doses of fertilizer were applied. Normal agronomic practices were followed for growing the crops. Each experimental unit was divided into two parts - one for sampling (termed as sampling area) with an area 14 m² and the other for estimating yield (termed as harvest area) with an area 10 m² . At dough stage (after panicle exertion), 8 hills were randomly selected from the sampling area and middle highest tiller from each selected hill were cut and taken to the laboratory for measuring five characters, namely, flag leaf angle (in^O), flag leaf area (cm²), all leaf area (cm²), weight of all leaves (gm) and dry weight of all leaves (gm). Days to 50% flowering for each variety were also recorded. At maturity, 10 hills were randomly selected from the sampling area, cut above the root level and taken to the laboratory for measuring seven characters viz. plant height (cm), number of effective tiller, panicle length for each tiller (cm), number of filled grain /hill, number of unfilled grain /hill, number of spikeletes/panicle and 1000 grain weight (gm). The matured crop was harvested from the harvest area separately for each plot and the yield was adjusted at 14% moisture content. Mahalanabis D² statistic was used for grouping the varieties. Mahalanobis D² statistic measured the distance between any two varieties and to allocate the varieties into sets of mutually exclusive groups such that varieties within a group are closer in terms of a prescribed set of characters while varieties in different groups are dissimilar. Canonical roots method was used to graphically represent the genotypic divergence and grouping the varieties into clusters. The method utilizes the multivariate variance-covariance structure of the varietal characteristics for computing two coordinates Z₁ and Z₂., which account for the largest share of the total phenotypic variance. From the graph, varieties are then grouped into distinct clusters based on their positions on the (Z₁, Z₂) plane.

Source of Information: BRRI Annual Report 2001-02, 97, BRRI, Gazipur.

Article Link (Online Journal):

Funding Source:

1. Government

Budget:

Total Budget (Tk.) :

Achievement/Findings:

Significant varietal differences were reflected for all characters on the significant F-values. Except for panicle length in T.Aman, all the F-values of the plant characters were significant at 1% level. Flag leaf angle was found to be the most variable plant character both in T Aman and in Boro as judged by CV. 1000 Grain weight, as expected, was the least variable character (CV =2.0% in T Aman and 2.4% in Boro) followed by plant height (CV= 2.8% in T Aman and 3.6% in Boro). Results showed that among the characters, 1000-grain weight, Plant height, total leaf area, panicle length, percent filled grain and number of filled grain/panicle are, in general, less affected by growing environment and may be assumed to be relatively stable compared to other plant characters.

Classification of genotype using Mahalanobis D²: The varieties were classified using Mahalanobis D² statistic. Six plant characters, namely, 1000 grain weight, plant height, panicle length, spikelets per panicle, flag leaf area and days to 50% flowering showed relatively stable contribution (%) to varietal divergence and were used for D² analysis. The stable contribution (%) of a character was obtained excluding the plant characters showing null or almost null contribution to the varietal divergence. The contribution (%) of plant character to the divergence of genotypes was the highest (69.5%) for days to 50% flowering in T Aman and was 65.8% in Boro (Table 8). 1000 Grain weight in T Aman and plant height in Boro ranked second accounting about the same proportion of contribution (23.8% and 22.6%, respectively) to the varietal divergence. The contribution of other characters was quite small. The results indicated days to 50% flowering as the most important character for grouping the varieties. The D² values of all pairs of 15 varieties in Boro (Table 9) and that for 20 varieties in T Aman showed significant results at least at 5% level. The results indicated that based on six characters, Mahalanobis D² distinguished the varieties exclusively. However, (BR10, BRRI dhan30) was the closest pair followed by the pairs (BR22, BRRI dhan30) and (BR10, BR23) among the T Aman varieties. In case of Boro varieties, BR2 was very much close to BR7 and BR8, and, BR18 was very much close to BR19.

Knowledge Management: Report/Proceedings