Juan Sebastian Panelo

The realized rate of genetic gain for grain yield (GY) and water productivity (WP) on-farm depends on the effectiveness of multi-environment trials (METs) to adequately sample the on-farm Target Population of Environments (TPE) and the distribution of on-farm gaps between current and potential yield as constrained by evapotranspiration (ET). Crop modeling permits simulation of genotype-by-environment-by-management combinations that can approximate biophysical limits for GY and WP for the TPE. However, the lack of scalable and cost-effective systems to estimate ET across region-wide METs prevent breeders from mapping their field trials onto the simulated “clouds of the possible” to make deliberate selection decisions for improved WP while selecting for GY. Here we demonstrate an approach that combines CERES-Sorghum, the satellite estimated ET available from OpenET, and GY from a 48 site-years MET. We show that sorghum breeding often operates at near biophysical boundaries, yet breeding gaps persist, creating opportunities to improve WP. Strong lack of correlation for GY among genotype-by-environment interactions could explain low rates of genetic gain for GY reported in grain sorghum. In the absence of deliberate efforts to improve WP, future yield improvement in sorghum will be inevitably tied to increases in water use and underpin the loss of drought tolerance. We conclude the scalable yield-ET gap-analysis methodology introduced here can enable breeders to enhance interpretation of breeding MET results and seek novel opportunities to increase both WP and GY, and that efforts to expand the application domain of the satellite estimated evapotranspiration beyond the U.S. are warranted.

Crop growth rate is a critical physiological trait for forage and bioenergy crops like sorghum [ Sorghum bicolor (L.) Moench], influencing overall crop productivity, particularly in photoperiod‐sensitive (PS) types. Crop growth rate studies focus on either a physiological approach utilizing a few genotypes to analyze biomass accumulation or a genetic approach characterizing easily scorable proxy traits in larger populations. Thus, the genetic control of crop growth rate in terms of biomass accumulation is poorly understood in PS sorghum. In this study, we monitored biomass accumulation in a diverse panel comprising 269 PS sorghum accessions in two growing seasons. We performed sequential samplings at 11 timepoints, separating leaves from stems. For the total biomass and each fraction, we applied the beta growth function to determine the maximum crop growth rate ( c _m ), maximum biomass accumulation ( w _max ), and time to c _m ( t _m ). Significant genetic variability was observed for all three parameters. Our analysis identified a practical window for c _m assessment through accumulated biomass at 60–70 days after planting. Genome‐wide association analysis suggested distinct and independent genetic controls of leaf and stem biomass accumulation, both physically and temporally. Common genomic regions were discovered controlling w _max and c _m of stem and total biomass. These results provide new insights into the genetic control of crop growth rate, highlighting promising genomic regions for functional validation. This research also offers practical applications for plant breeding programs demonstrating the feasibility of selecting superior genotypes for both early and late biomass accumulation to enhance crop productivity.

Canopy architecture traits are associated with productivity in sorghum [Sorghum bicolor (L.) Moench], and they are commonly measured at the time of flowering or harvest. Little is known about the dynamics of canopy architecture traits through the growing season. Utilizing the ground‐based high‐throughput phenotyping system Phenobot 1.0, we collected stereo images of a photoperiod‐sensitive and a photoperiod‐insensitive population over time to generate three‐dimensional (3D) representations of the canopy. Four descriptors were automatically extracted from the 3D point clouds: plot‐based plant height (PBPH), plot‐based plant width (PBPW), plant surface area (PSA), and convex hull volume (CHV). Additionally, genotypic growth rates were estimated for each canopy descriptor. Genome‐wide association analysis was performed on individual timepoints and the growth rates in both populations. We detected genotypic variation for each of the four canopy descriptors and their growth rates and discovered novel genomic regions associated with growth rates on chromosomes 1 (PBPH, CHV), 3 (PBPH), 4 (PBPH, PBPW), 5 (PBPH), 8 (PSA), and 9 (PBPW). These results provide new knowledge about the genetic control of canopy architecture, highlighting genomic regions that can be targeted in plant breeding programs.

Alternative physiological strategies can increase protein concentration in soybean: (i) more‐than‐proportional increases in seed protein content (mg seed⁻¹) relative to increases in carbohydrate and oil content in large‐seeded genotypes or (ii) more‐than‐proportional reductions in carbohydrate and/or oil content relative to protein content reductions in small‐seeded genotypes. Because these strategies differentially affect crop growth and development, we hypothesized that populations developed from high‐protein (HP) parents with contrasting seed sizes will present differences in how seed yield and protein concentration correlate. To test this, three breeding strategies were developed by mating high‐yielding cultivars and HP ones that differ in seed sizes, reflecting the alternative strategies mentioned above. Neither tested crossing strategies showed differences in their correlation values between seed yield and protein concentration, as initially expected. Nevertheless, populations developed from crossing a HP‐small‐seeded parent to a HP‐large‐seeded one showed the highest number of transgressive segregants for protein yield. Our results showed that parent selection based on seed size has no beneficial effects on the development of high‐yielding, HP soybean populations, but it might affect the number of transgressive segregants for protein yield.