Soybean is an important grain and oil crop worldwide and is rich in nutritional value. Phenotypic morphology plays an important role in the selection and breeding of excellent soybean varieties to achieve high yield. Nowadays, the mainstream manual phenotypic measurement has some problems such as strong subjectivity, high labor intensity and slow speed. To address the problems, a three-dimensional (3D) reconstruction method for soybean plants based on structure from motion (SFM) was proposed. First, the 3D point cloud of a soybean plant was reconstructed from multi-view images obtained by a smartphone based on the SFM algorithm. Second, low-pass filtering, Gaussian filtering, Ordinary Least Square (OLS) plane fitting, and Laplacian smoothing were used in fusion to automatically segment point cloud data, such as individual plants, stems, and leaves. Finally, Eleven morphological traits, such as plant height, minimum bounding box volume per plant, leaf projection area, leaf projection length and width, and leaf tilt information, were accurately and nondestructively measured by the proposed an algorithm for leaf phenotype measurement (LPM). Moreover, Support Vector Machine (SVM), Back Propagation Neural Network (BP), and Back Propagation Neural Network (GRNN) prediction models were established to predict and identify soybean plant varieties. The results indicated that, compared with the manual measurement, the root mean square error (RMSE) of plant height, leaf length, and leaf width were 0.9997, 0.2357, and 0.2666 cm, and the mean absolute percentage error (MAPE) were 2.7013%, 1.4706%, and 1.8669%, and the coefficients of determination (R2) were 0.9775, 0.9785, and 0.9487, respectively. The accuracy of predicting plant species according to the six leaf parameters was highest when using GRNN, reaching 0.9211, and the RMSE was 18.3263. Based on the phenotypic traits of plants, the differences between C3, 47-6 and W82 soybeans were analyzed genetically, and because C3 was an insect-resistant line, the trait parametes (minimum box volume per plant, number of leaves, minimum size of single leaf box, leaf projection area).The results show that the proposed method can effectively extract the 3D phenotypic structure information of soybean plants and leaves without loss which has the potential using ability in other plants with dense leaves.

Corn (Zea mays L.) is one of the key grain crops in China. In fields, the two crops of soybean (Glycine max L.) and peanut (Arachis hypogaea L.), which have nitrogen-fixing capacity (NFC), are generally used to intercrop with corn to improve plant physiology and production ability of corn even under fertilizer reduction. To explore a more scientific and reasonable way to plant corn, and simultaneously reduce the use of chemical fertilizers and pesticides, the impacts of corn intercropping with two NFC crops (including soybean and peanut) and the a non-NFC crop (i.e., millet (Setaria italica)) through five planting patterns, including three intercropping patterns (2 corn rows to 2, 3, and 4 NFC-crop rows or 2, 4, and 6 millet rows) and two sole crop patterns of corn and soybean, peanut, or millet under normal (600 kg/ha) and reduced (375 kg/ha) levels of NPK (N:P₂O₅:K₂O = 15:15:15) fertilization levels on the activity of N-metabolism-related enzymes in corn rhizosphere soil and corn leaves, and plant biomass and yield of corn were researched in this study. The results showed that fertilizer reduction significantly decreased the plant biomass and grain yield of the sole crop corn. The intercropping type and planting pattern both had significant effects on the activities of N-metabolism-related enzyme of soil alkaline protease (S-ALPT), and glutamine oxoglutarate aminotransferase (GOGAT), glutamate synthetase (GS), and nitrate reductase (NR) in the leaves of corn plants. The intercropping type of corn with soybean through the planting pattern of 2 corn rows to 4 soybean rows significantly improved the activities of N-metabolism-related enzymes in soil and corn leaves even under the fertilizer reduction. The intercropping pattern of corn-soybean was the most beneficial to increase the total nitrogen content in soil and corn leaves. In addition, the intercropping significantly increased the soil microbial diversity under normal fertilizer. Furthermore, fertilizer reduction significantly increased soil microbial diversity of the corn sole crop. Therefore, it is concluded that for corn in intercropping systems, the best and the worst companion crop were, respectively, soybean and millet.

Drought is one of the most important factors affecting plant growth and productivity. The previous results on drought tolerance (DT) genetic system in soybean indicated a complex of genes not only few ones were involved in the trait. This study is featured with a relatively thorough identification of QTL-allele/candidate-gene system using an efficient restricted two-stage multi-locus multi-allele genome-wide association study, on two comprehensive DT indicators, membership index values of relative plant weight (MPW) and height (MPH), instead of a single biological characteristic, in a large sample (564 accessions) of the Chinese cultivated soybean population (CCSP). Based on 24,694 multi-allele markers, 75 and 64 QTL with 261 and 207 alleles (2–12/locus) were detected for MPW and MPH, explaining 54.7% and 47.1% of phenotypic variance, respectively. The detected QTL-alleles were organized into a QTL-allele matrix for each indicator, indicating DT is a super-trait conferred by two (even more) QTL-allele systems of sub-traits. Each CCSP matrix was separated into landrace (LR) and released cultivar (RC) sub-matrices, which showed significant differentiation in QTL-allele constitutions, with 58 LR alleles excluded and 16 new ones emerged in RC. Using the matrices, optimal crosses with great DT transgressive recombinants were predicted. From the detected QTL, 177 candidate genes were annotated and validated with quantitative Real-time PCR, and grouped into nine categories, with ABA and stress responders as the major parts. The key point of the above results is the establishment of relatively full QTL-allele matrices composed of numerous gene functions jointly conferring DT, therefore, demonstrates the complexity of DT genetic system and potential of CCSP in DT breeding.

Seed weight and shape are important agronomic traits that affect soybean quality and yield. In the present study, we used image analysis software to evaluate 100-seed weight and seed shape traits (length, width, perimeter, projection area, length/width, and weight/projection area) of 155 novel recombinant inbred soybean lines (NJRISX) generated by crossing “Su88-M21” and “XYXHD”. We examined quantitative trait loci (QTLs) associated with the six traits (except seed weight per projection area), and identified 42 additive QTLs (5–8 QTLs per trait) accounting for 24.9–37.5% of the phenotypic variation (PV). Meanwhile, 2–4 epistatic QTL pairs per trait out of a total of 18 accounted for 2.5–7.2% of the PV; and unmapped minor QTLs accounted for the remaining 35.0–56.7% of the PV. A total of 28 additive and 11 epistatic QTL pairs were concentrated in nine joint QTL segments (JQSs), indicating that QTLs associated with seed weight and shape are closely related and interacted. An interaction was also detected between additive and epistatic QTL pairs and environment, which made significant contributions of 1.4–9.5% and 0.4–0.8% to the PV, respectively. We annotated 18 candidate genes in the nine JQSs, which were important for interpreting the close relationships among the six traits. These findings indicate that examining the interactions between closely related traits rather than only analyzing individual trait provides more useful insight into the genetic system of the interrelated traits for which there has been limited QTL information.