Previous PageTable Of ContentsNext Page

Selection of widely adapted lowland rice varieties for wet and dry seasons in Laos

Jaya Basnayake1,2, Shu Fukai2, Sipaseuth3, P Inthapanya3 and M Chanphengxay3

1BSES Limited, PO Box 117, Brandon 4807, Queensland, Australia
2
School of Land, Crop and Food Sciences, The University of Queensland, Brisbane 4072, Australia
3
National Agriculture and Forestry Research Institute, Vientiane, Lao PDR

Abstract

A series of lowland experiments, in the wet season (WS) under rainfed and dry season (DS) under irrigation, was conducted with 23 rice cultivars in four rice growing areas in Laos during the three-year period from 2003 WS to 2005/2006 DS. The objectives were a) to investigate the seasonal effects on cultivar performance in dry and wet season rice productivity in a range of rice growing environments in Laos and b) to identify the appropriate season for cultivar selection in the national rice breeding program. There was significant genotypic variation for grain yield (GY) only in four of ten experiments conducted in the WS. In contrast, genotypic variation for GY in all eight experiments conducted in the dry season was significant. The heritability (h2) for GY in WS experiments was lower (0.46) than the heritability for DS experiments (0.56). The genetic correlation between WS and DS experiments for grain yield was 0.36. While there was significant genotype-by-season interaction, the genotype clustering was not associated with the season and the results indicated that the increased use of DS screening would improve the efficiency of the overall rice variety selection program.

Key Words

rice, wet season, dry season, heritability, genotype-by-season interaction

Introduction

Most of the rice in the Lao PDR is cultivated in the WS. DS cultivation will become more important, as the rice requirement for local consumption is projected to increase by another 2 million tons by 2020. A recent increase in DS rice cultivation is associated with increased irrigation water availability. A series of experiment was conducted to investigate the consistency of variety performance between dry and wet seasons in a range of rice growing environments and to identify the best season for rice variety selection for wide adaptation in the breeding program.

Methods

Experimental sites

The experiments were conducted in WS and DS at four locations (L) during a three-year (Y) period from 2003 WS to 2005/2006 DS. The three locations were in Central and Southern Laos while one location was in the North. The North site was in Luang Namtha (LNT) at an altitude of 560 masl representing areas where early DS low temperature is a problem (Basnayake et al., 2006). The other three locations were the Agricultural Research Centre in Vientiane Municipality (VTN) (170 masl) in Northern Central Laos, Tassano station in the dry season and Seno District in the wet season at Savannakhet (SVK) in Southern Central Laos (110 masl) and Phonthong District in Champassak (CPK) (120 masl) in Southern Laos.

In total, 23 rice genotypes (G) from different phenological groups were evaluated. The early, medium and late maturing genotypes were tested together with widely adapted and popular varieties TDK1 and PNG3. The land preparation for the WS and DW experiments started in late May and early October, respectively. Cultivation practices typically used by farmers were adopted (Sipaseuth et al., 2007). Crops were established from transplanting with 30 days old seedlings and recommended fertilizer rates were applied at 3 stages of growth. Measurements were taken of the date of flowering (75% of the panicles emerged on primary culms), grain yield adjusted to 14% moisture content and spikelet sterility at harvest. Statistical analysis was conducted to estimate variance components for G, G×Y, G×L and G×Y×L for WS and DS experiments separately, and heritability (h2) estimated using variance components. The genetic correlation between WS and DS responses was estimated using the method of Kempthorne (1969). Cluster analysis (Cooper and DeLacy, 1994) was conducted with the G×E (WS and DS environment) matrix to identify patterns of environmental grouping for grain yield. Also the G, Season (S) and G×S interaction were analysed.

Results

Significant genotypic variation for grain yield was found only in four of the ten WS experiments. In contrast, genotypic variation for grain yield was significant in all eight DS experiments. Mean grain yield of DS and WS experiments at four locations are presented in Table 1. The DS yield was slightly higher than the WS yield (P<0.05).

Table 1. Mean grain yield (t ha-1) of 23 genotypes in wet and dry season at VTN, CPK, LNT and SVK.

Year

Season

VTN

CPK

LNT

SVK

Mean

04/05

Dry

3.27**

3.56**

3.14**

3.78**

3.45

05/06

Dry

3.76**

3.27**

2.08*

2.42*

2.88

Mean

 

3.52

3.42

2.61

3.10

3.17

2003

Wet

4.03NS

3.12**

4.50*

2.17NS

3.47

2004

Wet

2.89**

3.58NS

4.48NS

1.59NS

3.32

2005

Wet

-

2.50NS

-

1.89*

2.20

Mean

 

3.46

3.07

4.49

1.88

3.00

Overall mean

3.48

3.24

3.42

2.46

3.09

**, * and NS indicate genotypic variation significant at 1% and 5% levels and non- significant.

The estimated h2 for WS experiment was 0.46 (p<0.05) in contrast to 0.56 (p<0.05) for DS. Heritability for flowering dates in the DS was high and consistent across experiments. On average, DS experiments showed higher heritability for grain yield than WS experiments. There was a positive genetic correlation (0.36, p<0.05) between WS and DS experiments for the set of genotypes used. A significant genotype-by-season (G×S) interaction was identified in the combined analysis. However, the cluster analysis showed that environmental clustering was not highly associated with the season (Figure 1).

Figure 1. Hierarchical classifications of environments based on the genotype by environment interaction of 23 genotypes in 18 environments.

Conclusion

Genotypic variation was more strongly expressed in DS while environmental clustering was not associated with the season, indicating the advantage of DS screening to improve the efficiency of rice variety selection in Laos.

Acknowledgment

Financial assistance from ACIAR is greatly appreciated.

References

Basnayake J, Fukai, S, Sipaseuth, Schiller, JM., and Monthathip C (2006). Advances in Agronomic Research in the Lowland Rice Environments of Laos. In “Rice in Laos”. Ed. J.M. Schiller et al., Los Banos (Philippines), International Rice Research Institute, Manila Philippines. 349-369.

Cooper M. and Delacy IH (1994). Relationships among analytical methods used to study genotypic variation and genotype-by-environment interaction in plant breeding multi-environment experiments. Theoretical and Applied Genetics 88, 561–572.

Kempthorne O (1969). An Introduction to Genetic Statistics. The Iowa State University Press, Ames, Iowa, p. 545.

Sipaseuth, Basnayake J, Fukai S, Farrell TC, Senthonghae M, Sengkeo, Phamixay S, Linquist B, Chanphengsay M (2007). Opportunities to increasing dry season rice productivity in low temperature affected. Field Crops Research 102, 87-97.

Previous PageTop Of PageNext Page