Abstract In semiarid environments, early‐season water deficits create uncertainty in how much and when N fertilizer should be applied. To minimize indecision, a better understanding of the interrelationships among water availability, N rate, plant N sources, and yields is needed. This study investigated how soil water availability influenced maize ( Zea mays L.) yields and utilization of soil‐ and fertilizer‐derived N in semiarid environments in 2020 and 2021. The replicated trials were conducted at 12 sites and determined the impact of water availability and N rate on the sources of N within the grain, and precipitation and N use efficiencies (PUE and NUE, where NUE stands for nitrogen use efficiency and PUE stands for precipitation use efficiency). Based on grain δ 13 C and δ 15 N values, the relationship between water stress and fertilizer‐derived N was determined. At the three sites where maize was not responsive to N application, yields were limited by water availability, with an average yield of 5.7 Mg ha −1 . At the nine sites where maize responded to N application, (1) the average economic optimum N rate and yield were 137 kg N ha − 1 and 11.5 Mg ha −1 , respectively; (2) yield was highly correlated to plant available water (PAW) at planting + cumulative precipitation at tasseling (CP VT , r = 0.74, p < 0.05, where CP VT stands for cumulative precipitation at tasseling stage); and (3) soil‐derived N was positively correlated to PAW + CP VT ( r = 0.52, p < 0.01). The study also showed that increasing the N rate from 0 to 135 kg N ha −1 increased PUE by 39.1%, while increasing the N rate from 45 to 135 kg N ha −1 increased NUE by 13.4%.

IntroductionSoybean (Glycine max (L.) Merr.) is a multi-purpose crop, with yields that have increased significantly since 2000s in the United States. Given the complex mechanisms involved in vegetation growth, soybean yield estimation is determined by many factors, including weather, topographical attributes (TA), management practices, and cultivars. Conventional methods for estimating soybean crop maturity and yields are labor-intensive and include biases. Emerging methodologies for estimating crop maturity and yields that leverage remotely sensed data are increasingly gaining prominence. The objectives of this study were: (1) to identify the most relevant features derived from multi-source Earth observation data such as digital elevation model (DEM) based topography, vegetation indices (VIs), and soil indices (SIs)—for enhanced soybean yield estimation, and (2) to systematically evaluate the performance and robustness of extreme gradient boosting (XGBoost) model, in estimating on-farm soybean yields using “Leave a field out (LOFO) cross validation (CV)” approach.MethodTo achieve these objectives, we applied the XGBoost model to estimate soybean yields across three agricultural fields in South Dakota, USA, using data from the 2019 and 2021 growing seasons. The ML model estimated yields based on high-resolution PlanetScope satellite multispectral imagery and a DEM. The multispectral PlanetScope data were obtained for six soybean growth stages, ranging from early germination through maturity, and DEM was used to calculate the site’s topographic features. Soybean yield data were obtained from a GPS/yield montor equipped combine.ResultsThe XGBoost model had the highest R2 (0.54) at the Edmunds and Miner trial fields and the lowest RMSE (574 kg ha-1) and MAE (455 kg ha-1). A SHAP analysis showed that DVI exhibited the largest SHAP value range (approximately −1 to +1). Spatial maps showed differences in prediction error between Edmunds in 2019 and 2021. Over and underestimation are clearly visible through maps.DiscussionThese findings indicate that ML models can successfully estimate yields and underscore the importance of appropriate model selection. Accurate and timely on-farm yield estimation is a valuable tool that enables growers to make scientifically informed crop management and marketing decisions.

The purpose of this study was to examine adoption patterns of Uncrewed Aerial Vehicles (UAVs) in agriculture and to evaluate farmers’ perceived profitability of UAV use for remote sensing and input application. A statewide survey of 3,134 South Dakota farmers was conducted in 2025, generating 595 responses. Information was collected on UAV adoption, farm and operator characteristics, conservation practices, and perceived financial outcomes. To account for correlation between adoption decisions, a bivariate probit model was estimated to identify factors influencing the joint adoption of UAV imagery and UAV‑based input application. A parallel bivariate probit model was estimated to examine factors associated with perceived profit changes following UAV adoption. Nearly one quarter of respondents had adopted UAVs, with 18

A fall application of N fertilizer is believed to produce lower yields compared to a spring applied fertilizer due to excessive N loss. However, this assumption is based on research conducted in high rainfall areas that have very different N loss mechanisms than subhumid environments. Therefore, the objective was to evaluate the impact of a fall versus spring broadcast application of urea on corn (Zea mays L.) yields, the amount of soil and fertilizer derived N in the grain, inorganic N remaining in the soil, and leaching losses in a subhumid environment. The replicated research that was conducted in 2021, 2022, and 2023 measured (1) grain yields; (2) inorganic soil N in the fall, spring, and following harvest; and (3) the amount of N derived from soil and fertilizer using the N-15 natural abundance isotopic technique. The findings showed that fertilizer timing did not influence (1) corn grain yields (fall application produced 10.2 and spring application produced 10 Mg ha(-1)) across years, (2) the inorganic N remaining in the surface 60 cm after winter (84% of fall N was within the 30-cm depth), and (3) soil derived N percent in the grain (72% across years). These findings were attributed to the sorption of fall applied urea and they indicate that in this subhumid environment a broadcast application of urea in the fall (soil temperature <10 degrees C), to this well drained frigid soil, did not increase N losses or reduce yields.

Applying biochar and using cover crops are two potential approaches to reduce greenhouse gas (GHG) emissions. However, the effectiveness of these methods, individually or in combination, in salt-affected soils remains unclear. Thus, the objective was to determine the impact of barley (Hordeum vulgare) and biochar on N2O and CO2 emissions from salt-affected soil. During the 28-day replicated study, GHG emissions were measured near-continuously and the number of nirK, nirS, qnorB, and nosZ gene copies were measured 12 and 28 days after planting (DAP). Biochar accelerated barley emergence and reduced N2O–N and CO2–C emissions by 68% and 44% from 8 to 14 DAP, respectively. Barley reduced N2O–N emissions by 30.4% between 8 and 14 DAP, and at 12 DAP it reduced the number of nirK gene copies, that encodes for nitrite reductase by 40.9% and increased the number of nosZ gene copies, that encodes for nitrous oxide reductase by 193%. The biochar impact on N2O emissions was attributed to the 44% reduction in soil respiration, whereas the impact of barley was attributed to changes in the number of nirK and nosZ gene copies leading to increased efficiency of N2O reduction to N2. Overall, combining barley with biochar resulted in the greatest reduction (85%) of N2O emissions compared to soil alone.

Potassium fertilizer recommendations for optimal crop production may be improved by considering the ratio between expanding 2:1 layer silicates (smectite) with non-expanding 2:1 layer silicates (illite). However, the interactive effects between clay mineralogy and various soil tests are not well understood. This study evaluated the relationships among soil test K (STK), water-soluble K, HNO3-extractable K, pH, apparent cation exchange capacity (CECa), soil organic matter, clay content, and smectite:illite ratios. Soil samples (0-15 cm) were collected from 41 locations in central and eastern South Dakota, with textures ranging from sandy loam to clay. Data were partitioned into soils with smectite:illite ratios < 1 (illitic), >= 1 but <= 4.5 (smectitic), and > 4.5 (highly smectitic). Mean pH and CECa were lowest in illitic soils, higher in smectitic soils, and highest in highly smectitic soils, whereas STK and water-soluble K were lowest in highly smectitic soils relative to illitic and smectitic. Correlation analysis also showed that STK decreased with increasing smectite:illite ratio. These results suggest that exchangeable and water-soluble K forms are reduced when the proportion of smectite increases. There was a strong, positive relationship between STK and HNO3-extractable K across all three smectite:illite ratio groups. For soil pH, however, the relationship with STK was positive for illitic and smectitic soil groups, but negative for highly smectitic soils. Overall, these results suggest that the smectite:illite ratio influences the relationship among soil parameters and STK. This improves our understanding of the influence of clay mineralogy on plant-available K and the implications for K fertilizer recommendations.

Tillage intensity reduction when coupled with higher yields and better equipment, has increased the potential to sequester carbon in farm fields. However, a few experiments have demonstrated that this is occurring. This studies objective was to investigate the macro-scale effects of crop tillage intensity decreases and yield increases and on soil organic carbon (SOC) storage in Nebraska (NE), Iowa (IA), Minnesota (MN), and South Dakota (SD) from 2000 to 2021. The analysis was based on grower surveys, state yields from 2000 to 2021, and over 12 million surface soil samples that were aggregated by state and year. The model used first order kinetics, and it consisted of three pools [non-harvested carbon (NHC), SOC, and atmospheric carbon dioxide (CO2)]. Annual NHC additions were estimated from the state-level crop yields and tillage intensity reductions were estimated from producer surveys. Across the four states and 21 years, there was an estimated decrease of 0.0339 soil mixing events per year, corn (Zea mays) and soybean (Glycine max) yields increased by 63 and 38%, respectively, and SOC increased at a rate of > 460 kg SOC-C/(ha × year). In addition, strong (p < 0.01) linear correlations between NHC additions and SOC gains indicate that soil at the state-scale soil was not approaching carbon saturation.

Splitting N fertilizer application between the corn (Zea mays) plants pre-emergence and vegetative growth stage has the potential to reduce N losses while increasing N use efficiency. However, there are few published reports that show that splitting fertilizer reduces greenhouse gas emissions. Therefore, this study compared the CO2e (carbon dioxide equivalent) calculated from N2O-N and CO2-C emissions from a single pre-emergence application of 157 kg urea- N ha-1 with a nonfertilized control and two applications of 78.5 kg urea-N ha-1 that were applied at pre-emergence and at the corn plants V6 growth stage. Over three growing seasons (2021, 2022, and 2023), soil temperature, moisture, N2O-N and CO2-C emissions were measured six times per day from when the urea was applied to harvest. N2O-N and CO2-C emissions were separated into pre-split and post-split periods, and carbon dioxide equivalence was calculated. Splitting the N rate: (1) increased N2O-N emissions in 2021 and 2022; (2) reduced CO2-C emissions during the post-split period in 2022 and 2023; and (3) did not influence total CO2e in 2021 and reduced CO2e emissions in 2022 and 2023. Based on these results, splitting the N fertilizer rate between pre-emergence and the corn plants V6 growth stage should be considered as a potential climate smart practice.

This study aimed to estimate mixed-species cover crop (CC) biomass and nutrient contents using remote sensing, as ground-based measurements are time-consuming and costly. Eleven CC treatments with varying grass-legume proportions (GLP) were sampled, and nutrient contents were determined along with multispectral imagery captured during the first and fourth weeks of March and the fourth week of April 2023. Biomass N (R2 = 0.46-0.60) and K% (R2 = 0.41-0.71) decreased with increasing GLP. The chlorophyll absorption ratio index and the normalized difference vegetation index closely followed the biomass nutrients N, P, and K combined yield (Bio_NPK) trend. Machine learning algorithms random forest (RF) and partial least square (PLS) regression were better for biomass (R2 = 0.74 with RF) and N% (R2 = 0.72 with PLS) prediction compared to the Bio_NPK prediction. These results are crucial for scientists to devise appropriate analysis approaches for estimating the benefits of mixed-species CC.

Even though the impact of cover crops on soil moisture has been well documented, little work has investigated the resulting impact on methane emissions. Therefore, the objective was to determine the influence of a dormant seeded rye ( Secale cereale ) cover crop on soil temperatures, soil moisture, inorganic N, and total CH 4 ‐C emission in a well‐drained frigid soil from the start of growth in April/May through corn's ( Zea mays ) V4 growth stage. In this study, soil moisture, temperature, CH 4 ‐C, and N 2 O‐N fluxes were measured near‐continuously. The rye cover crop: (1) reduced the water‐filled porosity in the surface 5 cm in 2018 and 2020; (2) did not influence soil temperature prior to corn's V2 growth stage, increased the soil temperature between the V2 and V4 growth stages in 2019, and reduced the soil temperature in 2020; and (3) reduced ( p < 0.01) the CH 4 flux prior to corn seed emergence (VE) in 2018, 2019, and 2020. Over the 3 years, rye reduced ( p = 0.04) CH 4 flux (increased soil methane sink) by 6.4 g (ha × day) −1 prior to VE and reduced ( p = 0.02) carbon dioxide equivalence (CO 2 e) from −1.9 to −11.9 kg CO 2 e ha −1 . These decreases accounted for 22.3% of the CO 2 e derived from N 2 O (53.4 kg CO 2 e‐N 2 O) and suggest that the rye cover crop induced reductions in CH 4 emissions need to be considered in carbon intensity calculations.

Abstract The hypothesis of this research was that in a semiarid frigid environment, cover crop termination can be delayed beyond 2 weeks prior to planting without sacrificing corn (Zea mays) yield. To test this hypothesis, the impact of cover crop termination timing (2 weeks prior to, at corn planting, V2, and V4 corn) on cover crop biomass, soil water, inorganic N, and corn yields was quantified. Rye (Secale cereal) was dormant seeded in autumn (2018 and 2019), with corn planted the following May. In 2019 and 2020 growing seasons, precipitation was 607 and 324 mm, respectively, and corn growing degree days (base 10°C and maximum value of 30°C) were 1266 and 1436, respectively. Rye biomass increased when termination was delayed and averaged 25 kg ha−1 when terminated 2 weeks prior to corn planting and 911 kg ha−1 when terminated at V4. In both years, terminating the cover crop about 2 weeks prior to planting did not increase yields, but delaying termination from V2 to V4 reduced yields. In the dry year, when rainfall totaled 324 mm, delaying termination beyond corn planting (V0) reduced yields, whereas in the wet year (607 mm of rainfall) terminating rye at V2 did not reduce corn yields. These data suggest that in a semiarid environment, climatic conditions, especially early spring rainfall, must be carefully monitored when considering the cover crop termination date.

In agriculture, important unanswered questions about machine learning and artificial intelligence (ML/AI) include will ML/AI change how food is produced and will ML algorithms replace or partially replace farmers in the decision process. As ML/AI technologies become more accurate, they have the potential to improve profitability while reducing the impact of agriculture on the environment. However, despite these benefits, there are many adoption barriers including cost, and that farmers may be reluctant to adopt a decision tool they do not understand. The goal of this special issue is to discuss cutting-edge research on the use of ML/AI technologies in agriculture, barriers to the adoption of these technologies, and how technologies can affect our current workforce. The papers are separated into three sections: Machine Learning within Crops, Pasture, and Irrigation; Machine Learning in Predicting Crop Disease; and Society and Policy of Machine Learning. There are concerns that decision-making tools using machine learning and artificial intelligence (ML/AI) in farming could replace human knowledge and labor. ML/AI technologies have the potential to improve precision farming using site-specific algorithms. This special issue aims to open a dialogue about the role of ML/AI in agriculture and identify inequalities that may result from its use.

Because the manual counting of soybean (Glycine max) plants, pods, and seeds/pods is unsuitable for soybean yield predictions, alternative methods are desired. Therefore, the objective was to determine if satellite remote sensing-based artificial intelligence (AI) models could be used to predict soybean yield. In the study, multiple remote sensing-based AI models were developed for soybean growth stage ranging from VE/VC (plant emergence) to R6/R7 (full seed to beginning maturity). The ability of the deep neural network (DNN), support vector machine (SVM), random forest (RF), least absolute shrinkage and selection operator (LASSO), and AdaBoost to predict soybean yield, based on blue, green, red, and near-infrared reflectance data collected by the PlanetScope satellite at six growth stages, was determined. Remote sensing and soybean yield monitor data from three different fields in 2 years (2019 and 2021) were aggregated into 24,282 grid cells that had the dimensions of 10 m by 10 m. A comparison across models showed that the DNN outperformed the other models. Moreover, as crops matured from VE/VC to R4/R5, the R-2 value of the models increased from 0.26 to over 0.70. These findings indicate that remote sensing data collected at different growth stages can be combined for soybean yield predictions. Moreover, additional work needs to be conducted to assess the model's ability to predict soybean yield with vegetation indices (VIs) data for fields not used to train the model.

Our prior study showed that a dormant seeded rye (Secale cereale L.) cover crop reduced carbon dioxide equivalence (CO2e) in a frigid submesic climate. However, this research did not discuss the impact of the cover crop on temporal changes in soil inorganic nitrogen (N), soil organic carbon (SOC), microbial biomass, and community structure, which is the purpose of the present study. In this replicated no-tillage study, winter cereal rye was drilled in October of 2018 and 2019 and terminated using glyphosate (N-[phosphonomethyl] glycine) on June 11, 2019, and June 8, 2020, at the V4 growth stage in corn (Zea mays L.). Rye biomass collected at termination was air dried and placed into fiberglass litter bags at a rate of 3,750 kg ha-1. Bags were placed on the soil surface at the original collection zone. Four replicate bags from each plot were collected after 87, 248, and 365 days. Following removal, the material remaining in the removed bags was analyzed for amount and chemical composition. The decomposing cover crop biomass increased the microbial biomass and the amount of soil inorganic N and SOC in the surface 15 cm at 365 days (one year after litter was applied). These results suggest that soil nutrient benefits from fall-planted rye cover crop may not occur until the year following termination. However, the N credit benefits may vary depending on the type of cover crop, such as legume or brassica; therefore, this should also be considered in future studies. Overall, these findings provide practical implications for cover crop management in frigid submesic climates, suggesting that growers' adoption of cover crops can improve soil health, enhance crop yields, and reduce reliance on chemical fertilizers, thereby contributing to more sustainable and economically viable farming systems.

The slowly establishing salt-tolerant perennial grasses reduced nitrous oxide (N2O-N) emissions from saline/sodic soil compared to barren areas. Other salt-tolerant species may accelerate vegetative establishment and reduce N2O-N emissions. In a greenhouse study, barley (Hordeum vulgare L.), Florida broadleaf mustard (Brassica juncea L.), and Kernza intermediate wheatgrass [Thinopyrum intermedium (Host) Barkworth & D. R. Dewey] were grown for 63 days to compare shoot biomass and chemical composition, N2O-N emissions, and the soil microbiome between saline/sodic and productive (non-salt impacted) soils. Emissions were measured six times daily from 1 to 22 and 42 to 63 days after planting (DAP). Shoot and soil microbial biomass and communities were quantified 63 DAP. N2O-N emissions were 87% greater from no-plant saline/sodic than no-plant productive soil (p < 0.05). N2O-N emissions were reduced from planted treatments soon after plant emergence. N2O-N emissions reductions from saline/sodic soil during the first 22 DAP were 84%, 76%, and 61% for barley, mustard, and Kernza, respectively. Barley had the greatest shoot biomass and impact on the soil microbial community, increasing the fungi to bacteria ratio from 0.063 to 0.094 in the productive soil and from 0.056 to 0.076 in the saline/sodic soil. Plant-induced changes to the soil microbiome, and decreased soil inorganic N and water, contributed to N2O-N emission reductions. Archived field samples from grass-established saline/sodic soil areas had a fourfold increase in nos-Z gene copy number compared to no-plant controls, which may, in part, explain decreased N2O-N emissions. Establishing these vigorous species may aid in restoring multiple ecosystem services to saline/sodic areas.

Preharvest yield estimates can be used for harvest planning, marketing, and prescribing in-season fertilizer and pesticide applications. One approach that is being widely tested is the use of machine learning (ML) or artificial intelligence (AI) algorithms to estimate yields. However, one barrier to the adoption of this approach is that ML/AI algorithms behave as a black block. An alternative approach is to create an algorithm using Bayesian statistics. In Bayesian statistics, prior information is used to help create the algorithm. However, algorithms based on Bayesian statistics are not often computationally efficient. The objective of the current study was to compare the accuracy and computational efficiency of four Bayesian models that used different assumptions to reduce the execution time. In this paper, the Bayesian multiple linear regression (BLR), Bayesian spatial, Bayesian skewed spatial regression, and the Bayesian nearest neighbor Gaussian process (NNGP) models were compared with ML non-Bayesian random forest model. In this analysis, soybean (Glycine max) yields were the response variable (y), and spaced-based blue, green, red, and near-infrared reflectance that was measured with the PlanetScope satellite were the predictor (x). Among the models tested, the Bayesian (NNGP; R2-testing = 0.485) model, which captures the short-range correlation, outperformed the (BLR; R2-testing = 0.02), Bayesian spatial regression (SRM; R2-testing = 0.087), and Bayesian skewed spatial regression (sSRM; R2-testing = 0.236) models. However, associated with improved accuracy was an increase in run time from 534 s for the BLR model to 2047 s for the NNGP model. These data show that relatively accurate within-field yield estimates can be obtained without sacrificing computational efficiency and that the coefficients have biological meaning. However, all Bayesian models had lower R2 values and higher execution times than the random forest model. Remote sensing-based predictive models can estimate in-season soybean yields. Among the Bayesian models tested, accepting simplifying assumptions reduced the execution times and R2 values. All Bayesian models had lower R2 values and higher execution times than the non-Bayesian random forest model. Of the regression models, nearest neighbor Gaussian process has the highest R2 for training and testing data set. Model is transferable from one growing season to the next.

This paper analyzed farmers' willingness to accept (WTA) payment and uncertainties toward carbon market through a 2021 U.S. Midwest farmer survey. The findings showed that farmer carbon supply was elastic at intermediate prices ($20-50/Mg), but inelastic at low ($10-20/Mg) and high ($50-70/Mg) price levels. While perceived co-benefits play significant roles in promoting participation at low-price levels, variables such as age, education, farm size, and soil quality are more likely to influence producers' choices at intermediate- and high-price levels. To enhance farmers' support for carbon programs, measures should be taken to improve benefit, while reducing cost and uncertainty.

A growing body of research has examined farmers’ increasing economic challenges in the United States and the new models adopted to help them increase profit, remain in business, and achieve agricultural sustainability. However, the entrepreneurial strategies that Western Corn (Zea mays) Belt farmers use to overcome economic challenges and achieve agricultural sustainability remain understudied. The model system used in this study was eastern South Dakota, and it examined the entrepreneurial aspirations of commodity crop producers using mail and online survey data collected in 2018. Using the diffusion of innovations framework, we investigated how innovation and entrepreneurialism spread among farmers; whether frequent training, building, and using social networks were essential to farmers’ business success; and how age, education level, and farm size relate to their entrepreneurial aspirations. We analyzed these three socio-demographic characteristics of farmers against their adoption of entrepreneurship and engagement in networking and training. Our results show that (1) farmers are looking for ways to adopt entrepreneurship; (2) education and farm size are positively related to the adoption of entrepreneurship; (3) age is negatively related to farmers’ adoption of entrepreneurship, and (4) a larger farm size is associated with farmers’ use of social networks and their participation in training. This study highlights the importance of providing farmers with entrepreneurial training, equipping them with necessary skills, maximizing their use of social networks and opportunities, and encouraging strategic planning and best management practices.