The continuous assessment of cover crop growth throughout the season is a crucial baseline observation for making informed crop management decisions and sustainable farming operation. Precision agriculture techniques involving applications of sensors and unmanned aerial vehicles provide precise and prompt spectral and structural data, which allows for effective evaluation of cover crop biomass. Vegetation indices are widely used to quantify crop growth and biomass metrics. The objective of this study was to evaluate the accuracy of biomass estimation using a machine learning approach leveraging spectral and canopy height data acquired from unmanned aerial vehicles (UAVs), comparing different neural network architectures, optimizers, and activation functions. Field trials were carried out at two sites in Louisiana involving winter cover crops. The canopy height was estimated by subtracting the digital surface model taken at the time of peak growth of the cover crop from the data captured during a bare ground condition. When evaluated against the validation dataset, the neural network model facilitated with a Keras TensorFlow library with Adam optimizers and a sigmoid activation function performed the best, predicting cover crop biomass with an average of 96 g m⁻² root mean squared error (RMSE). Other statistical metrics including the Pearson correlation and R² also showed satisfactory conditions with this combination of hyperparameters. The observed cover crop biomass ranged from 290 to 1217 g m⁻². The present study findings highlight the merit of comprehensive analysis of cover crop traits using UAV remote sensing and machine learning involving realistic underpinning biophysical mechanisms, as our approach captured both horizontal (vegetation indices) and vertical (canopy height) aspects of plant growth.

High soil moisture due to frequent excessive precipitation can lead to reductions in maize grain yields and increased nitrogen (N) loss. The traditional methods of computing N status in crops are destructive and time-consuming, especially in waterlogged fields. Therefore, in this study, we used unmanned aerial vehicle (UAV) remote sensing to evaluate the status of maize under different N rates and excessive soil moisture conditions. The experiment was performed using a split plot design with four replications, with soil moisture conditions as main plots and different N rates as sub-plots. The artificial intelligence SciPy (version 1.5.2) optimization algorithm and spherical function were used to estimate the economically optimum N rate under the different treatments. The computed EONR for CRS 2022 was 157 kg N ha⁻¹ for both treatments, with the maximum net return to N of USD 1203 ha⁻¹. In 2023, the analysis suggested a lower maximum attainable yield in excessive water conditions, with EONR pushed up to 197 kg N ha⁻¹ as compared to 185 kg N ha⁻¹ in the control treatment, resulting in a lower maximum net return to N of USD 884 ha⁻¹ as compared to USD 1019 ha⁻¹ in the control treatment. This study reveals a slight reduction of the fraction of NDRE at EONR to maximum NDRE under excessive water conditions, highlighting the need for addressing such abiotic stress circumstances when arriving at an N rate recommendation based on an N-rich strip concept. This study confirms the importance of sensing technology for N monitoring in maize, particularly in supporting decision making in nutrient management under adverse weather conditions.

The rice water weevil, Lissorhoptrus oryzophilus (Coleoptera: Curculionidae), is the most damaging insect pest of U.S. rice. Larvae feed on rice roots greatly reducing plant health and yield. Quantifying infestations is made difficult by the subterranean feeding of L. oryzophilus larvae. A two-year field study was conducted to assess the feasibility of using remotely-sensed aerial imagery to quantify L. oryzophilus infestations and predict rice yields. Insecticidal seed treatments were used to establish rice plots with six levels of L. oryzophilus infestations in Louisiana during the 2019 and 2020 growing seasons. At three mid-season sampling times, unmanned aerial vehicles collected imagery of rice plots at heights of 10, 30, and 60 m. Images were then analyzed to obtain six different vegetation indices for each plot. Regression analysis assessed which sampling times, flight heights, and vegetation indices were best correlated with larval densities and rice yields. Numerous significant relationships were detected between vegetative indices and L. oryzophilus densities at all sampling dates at 30 and 60 m, but not at 10 m. Images taken approximately 3 weeks after flooding provided stronger correlations than images capture 4–5 weeks after flooding. Normalized Difference Vegetation Index, Green Normalized Difference Vegetation Index, Soil-Adjusted Vegetation Index, and Normalized Difference Red Edge repeatedly proved to be effective predictors of larval infestations. Visual Atmospheric Resistance Index and Modified Chlorophyll Absorption Reflectance Index produced weaker or unreliable correlations. All six indices were correlated with yields. Collectively, results demonstrated that L. oryzophilus infestations can be efficiently monitored through remotely-sensed aerial imagery.

Accelerating genetic gain in wheat breeding requires high-throughput phenotyping (HTP) to overcome labor-intensive bottlenecks and subjectivity associated with traditional field assessments. Unoccupied Aerial Systems (UAS) provide scalable, non-destructive phenotyping of crop development across breeding plots. Selecting optimal flight timing and predictive models is critical to ensure consistency across seasons. This study optimized UAS flight timing for wheat yield prediction by evaluating multi-year vegetation indices (VIs), including Normalized Difference Vegetation Index (NDVI), Normalized Difference Red Edge Index (NDRE), and Excess Green (ExG), aligned using Days Relative to Heading (DRH). Data from ten flights per year during the 2018–2021 growing seasons were used to assess predictive accuracy of machine learning models, including XGBoost, Ridge Regression, Random Forest, ExtraTrees, Partial Least Squares Regression, and Neural Networks. To address seasonal differences in planting date and sampling intervals, VIs were synchronized using DRH. Models were trained on 2018–2020 data and validated using an independent 2021 dataset. To enhance model explainability and ensure the reliability of the results, feature importance was evaluated over 100 iterations to identify the most consistent phenological window for yield prediction. XGBoost, ExtraTrees, and Random Forest achieved the lowest prediction errors (RMSE = 6.03 bu ac−1; NRMSE = 9.59%, RMSE = 6.1 bu ac-1; NRMSE = 9.7, RMSE = 6.58 bu ac-1, NRMSE = 10.47%) and showed negligible performance change in 2021 (ΔRMSE = -0.12 bu ac−1), indicating robust out-of-year generalization. VIs near heading were the most critical predictors, underscoring the importance of canopy condition during the transition to reproductive development. Concentrating UAS flights at heading therefore provides a resource-efficient approach for accurate yield prediction. Within the tested production environments, this framework supports high-throughput selection by identifying temporally stable canopy–yield relationships and optimizing UAS flight timing to reduce operational redundancy and cost.