Impact of genotype, seed composition, agronomic trait, and environment on soybean test weight

ABSTRACT

Test weight (TW) is a primary indicator of soybean grain quality and measure of bulk density (mass/volume). The objectives of this research were to evaluate the effects of genotype, environment and planting date on TW, and determine the relationship of TW with agronomic and seed composition traits. Six experiments (Exp) were conducted using diverse sets of cultivars/experimental lines. Significant negative linear correlations between TW and seed moisture in each of five genotypes were observed (P < 0.0001). A linear regression model using both measured TW and moisture were developed to normalize TW on a 13% moisture basis for comparison across genotypes and environments. Significant genetic variations (P < 0.01) of TW among genotypes were observed, ranging from 70.0 to 75.3; 70.3 to 74.3; and 66.4 to 74.3 kg hL⁻¹ in Exp-1, 2, and 3, respectively. Locations and G x E interactions also had significant effects on TW (P < 0.01) in these EXPs. Soybean yield was significantly negatively correlated with TW in Exp-1 and 3 (P < 0.01). Seed protein and sucrose contents, and maturity days were significantly, positively correlated with TW, while seed oil content and seed size were significantly, negatively correlated with TW in most or all experiments (P < 0.05). Effects of planting dates on TW were inconsistent for both Exp-4 and 5 between 2 years. There was no correlation between TW of the cultivars and their releasing years in MG VII. To our knowledge, this is the first report on test weight in soybean and the results could be used to guide future genetic improvement of TW.

KEYWORDS:

• Soybean
• test weight
• seed composition
• agronomic traits

Acknowledgments

Special thanks to Dale Wood, Brice Wilson, Earl Baxter, Greg Gokalp, and Tatyana Nienow for their technical assistance in this project and thanks Xuelin Luo for her technical assistance in data analysis.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the University of Georgia Research Foundation .