4.2. Phenotypic Data and Analysis

Fourier transform-near infrared reflectance (FT-NIR) spectrometry (Bruker, Karlsruhe, German) was applied to scan the near infrared absorption spectra of the dry seeds. Under the Quant 2 method of OPUS v. 4.2 software (Bruker, Karlsruhe, German), the samples’ protein and oil content data were calculated using the dry basis model [49]. Each RIL and soybean accession from each replication of each environment was detected three times using about 150~200 dry seeds per detection, with the average used in statistical analysis. The histogram of phenotypic data was constructed using EXCEL (Microsoft, Redmond, WA, USA). Statistical analysis of phenotypic data and ANOVA was conducted using SAS v. 9.4 (SAS Institute, Cary, NC, USA), with type III analysis being employed. The H² of protein and oil contents was calculated using the following equation [50]:

in which σG2 is genetic variance, σG∗E2 is genotype × environment variance, σe2 is error variance, r is the number of replications, and e is the number of environments.