Crop model determined mega-environments for cassava yield trials on paddy fields following rice

Abstract

The Cropping System Model (CSM)-MANIHOT-Cassava provides the opportunity to determine target environments for cassava (Manihot esculenta Crantz) yield trials by simulating growth and yield data for various environments. The aim of this research was to investigate whether cassava production on paddy fields in Northeast, Thailand could be grouped into mega-environments using the model. Simulations for four different cassava genotypes grown on paddy field following rice harvest was conducted for various soil types and the weather data from 1988 to 2017. The genotype main effect plus genotype by environment interaction (GGE biplot) technique was used to group the mega-environments. The analyses of yearly data showed inconsistent results across years for environment grouping and for the winning genotypes of the individual environment group. An analysis using GGE biplot with the average value of the simulated storage root dry weight (SDW) for 30 years indicated that all 41 environments were grouped into two different mega-environments. This study demonstrated the ability of the CSM-MANIHOT-Cassava to help identify the mega-environments for cassava yield trials on paddy field during off-season of rice that could help reduce both time and resources.

Keywords: Cropping system model; GGE biplot; Mega-environments; Multi-environment trials.

Fig. 1

Potential cassava production environments in the Northeast region of Thailand for upper paddy field during the off-season of rice.

Fig. 2

Relative contributions of environment (E), genotype (G), and G × E interaction effects expressed as percentage of total sum of squares (SS) in the analyses of variance for the years 1988–1989 to 2017–2018 for storage root yield (a) and total crop biomass (b).

Fig. 3

Genotype and genotype × environment biplots for four cassava genotypes simulation-tested across 41 environments for storage root yield in 1994–1995 (a) and in 2006–2007 (b) and total crop biomass in 1994–1995 (c) and in 2006–2007 (d) (E = environment; V1 = Kasetsart 50, V2 = Rayong 9, V3 = Rayong 11, and V4 = CMR38-125-77; PC = principal component).

Fig. 4

Genotype and genotype × environment biplot for mean of four cassava genotypes simulation-tested across 41 environments for storage root yields (a) and total crop biomass (b) from 1988 to 1989 to 2017–2018 (E = environment; V1 = Kasetsart 50, V2 = Rayong 9, V3 = Rayong 11, and V4 = CMR38-125-77; PC = principal component).