Seed Yield Prediction Research Articles

Prediction of Pea (Pisum sativum L.) Seeds Yield Using Artificial Neural Networks

A sufficiently early and accurate prediction can help to steer crop yields more consciously, resulting in food security, especially with an expanding world population. Additionally, prediction related to the possibility of reducing agricultural chemistry is very important in an era of climate change. This study analyzes the performance of pea (Pisum sativum L.) seed yield prediction by a linear (MLR) and non-linear (ANN) model. The study used meteorological, agronomic and phytophysical data from 2016–2020. The neural model (N2) generated highly accurate predictions of pea seed yield—the correlation coefficient was 0.936, and the RMS and MAPE errors were 0.443 and 7.976, respectively. The model significantly outperformed the multiple linear regression model (RS2), which had an RMS error of 6.401 and an MAPE error of 148.585. The sensitivity analysis carried out for the neural network showed that the characteristics with the greatest influence on the yield of pea seeds were the date of onset of maturity, the date of harvest, the total amount of rainfall and the mean air temperature.

Remote sensing for field pea yield estimation: A study of multi-scale data fusion approaches in phenomics.

Remote sensing using unmanned aerial systems (UAS) are prevalent for phenomics and precision agricultural applications. The high-resolution data for these applications can provide useful spectral characteristics of crops associated with performance traits such as seed yield. With the recent availability of high-resolution satellite imagery, there has been growing interest in using this technology for plot-scale remote sensing applications, particularly those related to breeding programs. This study compared the features extracted from high-resolution satellite and UAS multispectral imagery (visible and near-infrared) to predict the seed yield from two diverse plot-scale field pea yield trials (advanced breeding and variety testing) using the random forest model. The multi-modal (spectral and textural features) and multi-scale (satellite and UAS) data fusion approaches were evaluated to improve seed yield prediction accuracy across trials and time points. These approaches included both image fusion, such as pan-sharpening of satellite imagery with UAS imagery using intensity-hue-saturation transformation and additive wavelet luminance proportional approaches, and feature fusion, which involved integrating extracted spectral features. In addition, we also compared the image fusion approach to high-definition satellite data with a resolution of 0.15 m/pixel. The effectiveness of each approach was evaluated with data at both individual and combined time points. The major findings can be summarized as follows: (1) the inclusion of the texture features did not improve the model performance, (2) the performance of the model using spectral features from satellite imagery at its original resolution can provide similar results as UAS imagery, with variation depending on the field pea yield trial under study and the growth stage, (3) the model performance improved after applying multi-scale, multiple time point feature fusion, (4) the features extracted from the pan-sharpened satellite imagery using intensity-hue-saturation transformation (image fusion) showed better model performance than those with original satellite imagery or high definition imagery, and (5) the green normalized difference vegetation index and transformed triangular vegetation index were identified as key features contributing to high model performance across trials and time points. These findings demonstrate the potential of high-resolution satellite imagery and data fusion approaches for plot-scale phenomics applications.

Site-Specific Seed Yield Prediction of Red Fescue (Festuca rubra L.) Based on Drone Imaging and Local Regression Models

Yield maps give farmers information about growth conditions and can be a tool for site-specific crop management. Combine harvesters may provide farmers with detailed yield maps if there is a constant flow of a certain amount of biomass through the yield sensor. This is unachievable for grass seeds because the weight of the intake is generally too small to record the variation. Therefore, there is a need to find another way to make grass seed yield maps. We studied seed yield variation in two red fescue (Festuca rubra) fields with variation in management and soil fertility, respectively. We estimated five vegetation indices (VI) based on RGB images taken from a drone to describe yield variation, and trained prediction models based on relatively few harvested plots. Only results from the VI showing the strongest correlation between the index and the yield are presented (Normalized Excess Green Index (ExG) and Normalized Green/Red Difference Index (NGRDI)). The study indicates that it is possible to predict the yield variation in a grass field based on relatively few harvested plots, provided the plots represent contrasting yield levels. The prediction errors in yield (RMSE) ranged from 171 kg ha−1 to 231 kg ha−1, with no clear influence of the size of the training data set. Using random selection of plots instead of selecting plots representing contrasting yield levels resulted in slightly better predictions when evaluated on an average of ten random selections. However, using random selection of plots came with a risk of poor predictions due to the occasional lack of correlation between yield and VI. The exact timing of unmanned aerial vehicles (UAVs) image capture showed to be unimportant in the weeks before harvest.

High-Resolution Flowering Index for Canola Yield Modelling

Canola (Brassica napus), with its prominent yellow flowers, has unique spectral characteristics and necessitates special spectral indices to quantify the flowers. This study investigated four spectral indices for high-resolution RGB images for segmenting yellow flower pixels. The study compared vegetation indices to digitally quantify canola flower area to develop a seed yield prediction model. A small plot (2.75 m × 6 m) experiment was conducted at Kernen Research Farm, Saskatoon, where canola was grown under six row spacings and eight seeding rates with four replicates (192 plots). The flower canopy reflectance was imaged using a high-resolution (0.15 cm ground sampling distance) 100 MP iXU 1000 RGB sensor mounted on an unpiloted aerial vehicle (UAV). The spectral indices were evaluated for their efficiency in identifying canola flower pixels using linear discriminant analysis (LDA). Digitized flower pixel area was used as a predictor of seed yield to develop four models. Seventy percent of the data were used for model training and 30% for testing. Models were compared using performance metrics: coefficient of determination (R2) and root mean squared error (RMSE). The High-resolution Flowering Index (HrFI), a new flower index proposed in this study, was identified as the most accurate in detecting flower pixels, especially in high-resolution imagery containing within-canopy shadow pixels. There were strong, positive associations between digitized flower area and canola seed yield with the peak flowering timing having a greater R2 (0.82) compared to early flowering (0.72). Cumulative flower pixel area predicted 75% of yield. Our results indicate that the HrFI and Modified Yellowness Index (MYI) were better predictors of canola yield compared to the NDYI and RBNI (Red Blue Normalizing Index) as they were able to discriminate between canola petals and within-canopy shadows. We suggest further studies to evaluate the performance of the HrFI and MYI vegetation indices using medium-resolution UAV and satellite imagery.

Odnos komponenata prinosa i prinosa zrna soje (Glycine max L. Merr.) u uvjetima navodnjavanja

The study presents the results of a three-year experiment (2013–2015) that was carried out to determine a relationship between the soybean yield components and the seed yield under different irrigation treatments. The results indicated that the study year had the greatest effect on the number of nodes per plant (NNP), while an interaction between the irrigation and experiment year was also statistically significant. The highest average NNP was observed in 2015, being 33% higher when compared to the year 2014. The highest number of seeds per plant (NSP) was observed in 2015, being 20% and 31% higher when compared to 2013 and 2014. An abundant irrigation resulted in the highest NSP when compared to a rational and control treatment. Irrigation, study year, and their interaction did not have a statistically significant effect on the thousand seed weight (TSW) (g), but the lowest average TSW (g) was obtained in the control treatment of each study year. Regression models pertaining to the seed yield prediction in the control treatment and rational irrigation were not statistically significant. However, in the abundant irrigation, the regression model based on the TSW (g), NNP, and the NSP as the predictors provided for a statistically significant model seed yield prediction, but only the NSP was identified as a highly significant seed yield predictor.

Impact of winter oilseed rape nutritional status during vegetative growth on yield

The nutritional status of winter oilseed rape (WOSR) during its vegetative period is crucial for plant growth and can be used for the seed yield prediction. This hypothesis was verified based on the data from long-term field experiments. The experiment consisted of four potassium (K) treatments based on the progressive K supply potential to plants from soil and fertilizer and two magnesium treatments (–Mg, +Mg) conducted in 2013–2015. The content of nutrients (N<sub>tot</sub>, P, K, Mg, Ca, Fe, Mn, Zn, Cu) was determined at the rosette stage (BBCH 30) for leaves and separately for leaves and stems in the late stage of inflorescence growth (BBCH 57–59). The low K content appeared as the key limiting nutrient in WOSR plants in the rosette stage due to the insufficient soil fertility level, depended even more on weather conditions. This negative K nutritional trait persisted through the whole vegetative WOSR growth. Its detection was possible, because stems were included in the diagnostic procedure. The most reliable prognosis of WOSR yield was conducted based on the nutritional status of stems in the late stage of the inflorescence development.

Estimation of Sunflower Seed Yield Using Partial Least Squares Regression and Artificial Neural Network Models

Statistical models can efficiently establish the relationships between crop growth and environmental conditions while explicitly quantifying uncertainties. This study aimed to test the efficiency of statistical models established using partial least squares regression (PLSR) and artificial neural network (ANN) in predicting seed yields of sunflower (Helianthus annuus). Two-year field trial data on sunflower growth under different salinity levels and nitrogen (N) application rates in the Yichang Experimental Station in Hetao Irrigation District, Inner Mongolia, China, were used to calibrate and validate the statistical models. The variable importance in projection score was calculated in order to select the sensitive crop indices for seed yield prediction. We found that when the most sensitive indices were used as inputs for seed yield estimation, the PLSR could attain a comparable accuracy (root mean square error (RMSE) = 0.93 t ha−1, coefficient of determination (R2) = 0.69) to that when using all measured indices (RMSE = 0.81 t ha−1, R2 = 0.77). The ANN model outperformed the PLSR for yield prediction with different combinations of inputs of both microplots and field data. The results indicated that sunflower seed yield could be reasonably estimated by using a small number of crop characteristic indices under complex environmental conditions and management options (e.g., saline soils and N application). Since leaf area index and plant height were found to be the most sensitive crop indices for sunflower seed yield prediction, remotely sensed data and the ANN model may be joined for regional crop yield simulation.

Incorporating pleiotropic quantitative trait loci in dissection of complex traits: seed yield in rapeseed as an example

Key messageA comprehensive linkage atlas for seed yield in rapeseed.Most agronomic traits of interest for crop improvement (including seed yield) are highly complex quantitative traits controlled by numerous genetic loci, which brings challenges for comprehensively capturing associated markers/genes. We propose that multiple trait interactions underlie complex traits such as seed yield, and that considering these component traits and their interactions can dissect individual quantitative trait loci (QTL) effects more effectively and improve yield predictions. Using a segregating rapeseed (Brassica napus) population, we analyzed a large set of trait data generated in 19 independent experiments to investigate correlations between seed yield and other complex traits, and further identified QTL in this population with a SNP-based genetic bin map. A total of 1904 consensus QTL accounting for 22 traits, including 80 QTL directly affecting seed yield, were anchored to the B. napus reference sequence. Through trait association analysis and QTL meta-analysis, we identified a total of 525 indivisible QTL that either directly or indirectly contributed to seed yield, of which 295 QTL were detected across multiple environments. A majority (81.5%) of the 525 QTL were pleiotropic. By considering associations between traits, we identified 25 yield-related QTL previously ignored due to contrasting genetic effects, as well as 31 QTL with minor complementary effects. Implementation of the 525 QTL in genomic prediction models improved seed yield prediction accuracy. Dissecting the genetic and phenotypic interrelationships underlying complex quantitative traits using this method will provide valuable insights for genomics-based crop improvement.

Seed yield and protein content in the Weibullsholm Pisum collection

In Europe, agriculture is highly dependent on imported soybean from South America. Potential alternative sources are protein from peas (Pisum sativum L.) or more local sources like other grain legumes or rapeseed meal (Brassica napus L. subsp. oliefera). These are also good rotation crops. For farmers, protein and yield are key traits. In this study, a dataset containing 37 descriptors and 1222 accessions from a germplasm collection of P. sativum was analyzed. Scatterplot matrixes and tree regression analysis were used to establish the relationship among descriptors and to identify the most important predictors for seed yield and protein content respectively. Number of flowers per plant was shown to be important for seed yield prediction, followed by number of inflorescences per plant and number of pods per plant. In general, a negative correlation between seed protein content and seed yield was detected, but a few accessions that had both high seed yield and high protein content were identified. The results are discussed in relation to crop improvement and the importance of maintaining germplasm collections.

Sunflower seed yield estimation under the interaction of soil salinity and nitrogen application

Understanding the interaction effects of salinity and fertilizer is of great economic and environmental importance for improving crop yield and fertilizer used efficiency. The present study established a sunflower seed yield prediction model for saline soils directly from soil salinity (S) and nitrogen application (N) and related the results to the optimum N rate in saline soils. Two years of field trial data on sunflower growth under different S levels and N rates were collected at the Yichang Experimental Station in the Hetao Irrigation District of Inner Mongolia, China. Statistical techniques, including bootstrap resampling and artificial neural network (ANN) analysis, were used for data analysis and model establishment. The results indicated that sunflower crop characteristics varied in relation to S and N rates. The changes in plant height (H) with time before the late mature stage could be fitted with a linear function under different S and N treatments. Furthermore, the leaf area index (LAI) varied from the day after sowing and was expressed using a quadratic polynomial. The salt tolerance threshold (ECth) values for sunflower at the seedling, bud, flowering, and mature stages were 6.93, 14.46, 14.46, and 17.26dSm−1, respectively. When S was smaller than ECth, relative evapotranspiration (RET) fluctuated with S and was estimated using an artificial neural network (ANN); moreover, when S was larger than ECth, RET decreased with S and was simulated using linear regression. In addition, effective nitrogen application (ENA) was determined coupled with S, and the positive and negative effects of ENA on RET were expressed as a quadratic polynomial. The sunflower seed yield was estimated using a modified Jensen model (SNJ model) (R2=0.51). The maximum relative seed yield (MRSY) decreased rapidly with soil salinity until 12dSm−1; when the soil salinity level was below 18dSm−1, the optimal N application rate fluctuated from 100 to 160kgha−1.

Stepwise regression and principal component analyses for quantitative traits of rapeseed genotypes at different sowing dates

The present research was done to assess the best selection criteria for yield improvement in rapeseed (Brassica napus L.) using stepwise regression and principal component analyses at different sowing dates. All the traits except 1000-seed weight were significantly affected by sowing dates. The results of stepwise regression analysis revealed that seeds per pod had an important role at the first and second sowing dates, but at the third and fourth sowing dates, pods per plant and days to flowering were more important than other yield components for a seed yield prediction model. On the basis of a cumulative percent of variation, three principal components (PCs) were determined for each sowing date. The cumulative percentages of variation for three PCs at the first to fourth sowing dates were 0.97, 0.96, 0.89 and 0.95, respectively. At the first sowing date, the first principal component (PC1) had high positive and negative PC loading values for the studied traits such as days to flowering, days to the end of flowering, duration of flowering, pods per plant and harvest index. Therefore, there was high variation in these traits among the genotypes. PC2 of the first sowing date had also high PC loadings for pods on the main raceme, seeds per pod, 1000-seed weight, biological and seed yields, therefore the correlation of these traits with this PC will be high. In PC3 of the first sowing date, height, pods on the main raceme and pods per plant had the high value of PC loadings. Based on stepwise regression analysis, seeds per pod at the first and second sowing dates and days to flowering and pods per plant at the third and fourth sowing dates had an important role for improving seed yield.

Coefficient Analysis and Association between Morpho-Agronomical Characters in Common Bean (Phaseolus vulgaris L.)

AbstractPulses are important food sources for human consumption. In an attempt to study the associations between common bean (Phaseolus vulgaris L.) morpho-agronomical traits, twelve genotypes and inbred lines were evaluated in a randomized complete block design (RCBD) with three replications in 2010-2011 crop season. Seed yield and 20 other morpho-agronomic characters were recorded. Analysis of variance (ANOVA) revealed a significant (p ≤ 0.01) difference among genotypes with respect to almost all traits. Correlation analysis demonstrated that seed yield had a strong positive correlation with both seed number per plant and seed number per pod, suggesting the usefulness of these traits in common bean breeding programs. Step-wise regression analysis pointed out that the pod weight, seed number per pod and 100 seed weight contributed to the seed yield prediction, whereas other traits did not contributed to the seed yield prediction. These traits explained almost 99% of total seed yield variations. Path analysis showed that the maximum direct and positive effect was related to pod weight. Furthermore, factor analysis revealed that four factors, explained almost 71% of the total variance. The results of this research showed that biologic yield, pod weight, straw weight, number of pod per plant and seed number per plant were the most closely related traits to the seed yield.

Seed yield prediction of sesame using artificial neural network

The prediction of seed yield is one of the most important breeding objectives in agricultural research. So, in this study, two methods namely artificial neural network (ANN) and multiple regression model (MLR) were employed to estimate the seed yield of sesame (SYS) from readily measurable plant characters (e.g., flowering time of 100% (days), the plant height (cm), the capsule number per plant, the 1000-seed weight (g) and the seed number per capsule). The ANN and MLR were tested using field data. Results showed that the ANN predicts the SYS accurately with a root-mean-square-error (RMSE) of 0.339t/ha and a determination coefficient (R2) of 0.901. Also, it was found that the ANN model performed better than the MLR model with a RMSE of 0.346t/ha, and R2 of 0.779. Finally, sensitivity analysis was conducted to determine the most and the least influential characters affecting SYS. It was found that the capsule number per plant and the flowering time of 100% had the most and least significant effects on SYS, respectively.

A conceptual framework for Jatropha seed yield estimation using adaptive neuro-fuzzy inference system (ANFIS) modelling

Jatropha seed yield prediction is one of the most important influencing factors for developing a supply chain modelling of Jatropha seed. The oil from this Jatropha seed is used to blend with diesel to obtain biofuel (Hiromi, Yamamoto, Junichi Fujino, and Kenji Yamaji. 2001. “Evaluation of Bioenergy Potential with a Multi-regional Global-Land-Use-and-Energy Model.” Biomass and Bioenergy 21: 185–203). The Jatropha plant is easy to cultivate and produces high yield if properly maintained. The main focus of this study is to utilize uncultivable wastelands for cultivating Jatropha seeds. As suggested by Liu et al. (Liu, Xiaohong, David B. Grant, Alan C. McKinnon, and Yuanhua Feng. 2010. “An Empirical Examination of the Contribution of Capabilities to the Competitiveness of Logistics Service Providers: A Perspective from China.” International Journal of Physical Distribution & Logistics Management 40 (10): 847–866), the effectiveness of prediction affects the functional characteristics of a supply chain network design. The yield prediction of Jatropha seeds has two important roles which include (i) the identification of external parameters that affects the yield and (ii) the detection of internal attributes that changes the growth characteristics of the Jatropha plant. The development of the fuzzy inference system is characterized by a large number of input variables (Dobrila Petrovic. 2001. International Journal of Production Economics 71: 429–438). A Matlab programming software was used to integrate an adaptive neuro-fuzzy inference system. This approach gave the numerical as well as graphical output that was used to interpret the final result. The root mean square error values were identified for the given inputs which were then compared with the trained input variables to select the best input among the given alternative variables.

Impact of weather variables on yield and yield attributes in Okra under different growing environments

Okra variety Hisar Unnat was sown on eight different dates during March to July of 2008 and 2009. Results revealed that high seed yield was obtained in okra sown on 15th June, 25th June and 5th July Seeds per fruit, test weight and standard germination were also found superior during this period. Average relative humidity had significantly positive correlation with seed yield (r=0.76), number of seeds per fruit (r=0.72), test weight (r=0.79) and standard germination (r=0.79). Maximum, minimum and mean temperatures and evaporation expressed significantly negative association with above parameters. High seed yield of okra can be harvested, when maximum, minimum, and mean temperature remains around 33.5-36oC, 19.5-24oC, and 26.9-30oC, respectively with average relative humidity 65-75% and 4.1 -5.7 mm evaporation. Seed yield prediction would be more reliable when temperature, relative humidity and evaporation were taken together instead of individual weather variable in regression equation.

Seed biomass production and germination rates of Cyperus entrerianus1

Abstract Cyperus entrerianus (Boeckler) (deeprooted sedge, deeproot sedge, woodrush flatsedge) has become naturalized throughout the Texas coastal prairie ecoregion and is now found throughout much of the southeastern United States. Originally from Argentina, its successful naturalization is linked to its rapid growth, morphological plasticity, and generalized habitat requirements, but little work has focused upon its seed biology or ecology. Research was conducted to quantify phytomorphological variability, develop seed yield prediction models, and quantify germination requirements for C. entrerianus to aid in future control and management efforts of this exotic invasive sedge, particularly those focused upon seedbanks. Cyperus entrerianus phytomorphology, seed production, and seed germination were extremely variable among study areas, and provide evidence for its successful naturalization. Although plants varied morphologically, spike cluster area and inflorescence diameter explained much of the variati...

Field Evaluations of Yield, Iron-Manganese Relationship, and Chlorophyll Meter Readings in Soybean Genotypes as Affected by Iron-Ethylenediamine Di-o-hydroxyphenylacetic Acid in a Calcareous Soil

ABSTRACT Iron (Fe)-deficiency chlorosis is a common constraint when soybean (Glycine max L.) is grown on calcareous soils. Considerable differences exist among soybean genotypes for susceptibility to Fe chlorosis. In order to evaluate the effectiveness of iron-ethylenediamine di-o-hydroxyphenylacetic acid (Fe-EDDHA) for three soybean genotypes (A3237, Black hack, and Wells), field studies were conducted for the years 2001 and 2002 in a calcareous soil. Although, available Fe of the studied soils was either lower than critical level or in marginal range, application of Fe-EDDHA did not result in a significant increase in soybean yield probably, due to the antagonistic relationships between Fe and manganese (Mn). It appears that Fe soil test as the only criterion for Fe fertilizers recommendation is not appropriate and soil test for Mn is also recommended. Significant quadratic equations were obtained between chlorophyll meter readings (CMR) in growth stage 3 (GS3) with seed yield (SY) of A3237 and Black hack. However, SY of Wells showed close relationships with CMR in growth stage 4 (GS4). A tentative conclusion is that the chlorophyll meter is a reliable and non-destructive tool for the prediction of SY for the studied soybean genotypes in GS3 or GS4. Due to the fact that use of Fe fertilizer might cause nutritional disorder, use of Fe-efficient genotypes remains as an effective and economic sound practice. However, positive responses to Fe-chelate treatments are expected for genotypes with shoot Fe: Mn ratio less than 0.4 when both nutrients are in the sufficiency range. Obviously, such a requirement limits the utilization of Fe-chelate to post-emergence fertilization.

Adaptive extent and seed yield predictions for Microula sikkimensis grown in the Qinghai-Tibet Plateau, China

Microula sikkimensis (Family Boraginaceae) is a species with multiple potential applications in medicine, food and fodder. From 1994 to 2004, field surveys and transplanting trials were conducted in eleven counties of three provinces in China in order to collect information about the habitat of this species. Extending Hutchinson's concept of a niche as an n-dimensional super-volume, the habitat niche-fitness (HNF) of M. sikkimensis is defined as the degree of similarity between the supply of an actual habitat and the requirement for the habitat to be optimum, in which the supply of the actual habitat and the requirement for the optimum habitat represent realistic habitat conditions and species attributes, respectively. A new model of HNF is constructed to evaluate the adaptive extent of M. sikkimensi with the main habitat factors as independent variables and factor weights as parameters. The calculation results show that the model of HNF notably outperforms the proportional similarity index and the geometric parallelism formula both in mathematical justification and biological principle testing. With HNF as a surrogate for composite environmental factors, a seed yield prediction model was selected from all probable regression models by cross validation method. The apparent prediction error and 45° line test showed that there was no significant difference between the observed seed yield from the test set and the corresponding expected seed yield values from the selected model, confirming the validity of the selected model in predicting seed yield across a wide-spread area. The seed yield prediction model and its uses are recommended within the limitations of the data used in the study area. Beyond this range, validation of the seed yield prediction model will be necessary.

Traits associated with dry edible bean ( Phaseolus vulgaris L.) productivity under diverse soil moisture environments

Two experiments were conducted in the Rift Valley, Ethiopia (8°N and 39°E) to determine associations between eight plant traits and seed yield, and to obtain estimates of narrow sense heritability for the traits. Experiment I evaluated seven dry edible bean cultivars/lines at two locations to simulate different soil moisture stress, including, Debre Zeit(non-stress) and Dera (moderate-stress). Experiment II evaluated 25 cultivars/lines in three environments including, Melkassa early planted (non-stress), Melkassa late planted (high-stress), and Dera (moderate-stress). A randomized-complete-block design with three replicates was used in both experiments. Plant traits evaluated were seed yield, pods plant-1, seeds pod-1, 100 seed weight, root dry weight, hypocotyl diameter, plant biomass, plant height and days to flowering. Plant traits that were significantly associated with seed yield were included in a stepwise-regression model to determine which trait or combination of traits provided the best model to estimate seed yield in each environment. An analysis of variance was conducted to test main effects and interactions between plant traits and environments. Significant variation among lines occurred for seed yield and all plant traits in both experiments. Strong positive correlations were observed between plant biomass and seed yield in all environments. Seed yield and pods plant-1 were also highly associated in four of the five environments. Stepwise regression models indicated that the combination of pods plant-1 and plant biomass consistently contributed to seed yield prediction, while other traits did not. Because both plant biomass and pods plant-1 had moderate to high narrow sense heritability estimates and low GE interactions, they should be useful as indirect selection criteria to improve and stabilize seed yield in a breeding program.