Springer Science and Business Media LLC

This study analyzed the spatial distribution of Yellow Sigatoka Leaf Spot relative to soil fertility and plant nutritional status using geostatistics. The experimental area comprised 1.2 ha, where 27 points were georeferenced and spaced on a regular grid 18 × 18 m. The severity of Yellow Sigatoka, soil fertility and plant nutritional status were evaluated at each point. The spherical model was adjusted for all variables using restricted maximum likelihood. Kriging maps showed the highest infection rate of Sigatoka occurred in high areas of the field which had the highest concentration of sand, while the lowest disease was found in lower areas with lower silt, organic matter, total exchangeable bases, effective cation exchange capacity, base saturation, Ca and Mg in soil, and foliar sulfur (S). These results may help farmers manage Yellow Sigatoka disease more effectively, with balanced fertilization and reduced fungicide application. This practice minimizes the environmental impact and cost of production while contributing to production sustainability.

A procedure for estimating the number of mature apples in orchard images captured at night-time with artificial illumination was developed and its potential for estimating yield was investigated. The procedure was tested using four datasets totaling more than 800 images taken with cameras positioned at three heights. The procedure for detecting apples was based on the observation that the light distribution on apples follows a simple pattern in which the perceived light intensity decreases with the distance from a local maximum due to specular reflection. Accordingly, apple detection was achieved by detecting concentric circles (or parts of circles) in binary images obtained via threshold operations. For each dataset, after calibration of the procedure using 12 images, the estimates of the number of apples were within a few percent of the number of apples counted by visual inspection. Yield estimations were obtained via multi-linear models that used between two and six images per tree. The results obtained using all three cameras were only slightly better than those obtained using only two cameras. Using images from only one side of the tree did not worsen the results significantly. Overall, the yield estimated by the best models was within $$\pm$$ 10 % of the actual yield. However, the standard deviation of the yield estimation errors corresponded to ~26–37 % of the average tree yield, indicating that improvements are still needed in order to achieve accurate yield estimation at the single-tree level.

On-farm adoption of individual and groups of precision agriculture technologies has grown in the past 15 years. Based on a sample of 545 farm observations collected by the Kansas Farm Management Association, farm adoption of bundles of embodied knowledge and information intensive technologies was analyzed using a Markov transition approach. Three separate analyses estimated transition probabilities to show the adoption of bundles of embodied knowledge technologies, the adoption of bundles of information intensive technologies, and the adoption of variable rate technologies contingent on prior adoption of embodied knowledge and/or information intensive technologies. Each analysis was estimated for two separate time periods (2009–2012) and (2013–2016). The probability that farms retain the same bundle or transition to a different bundle by the next time period are reported. The results indicate that persistence with the same technology bundle is the predominant behavior and that this behavior has strengthened in the study’s most recent time period.

Commercial consultants frequently sell site-specific crop input management recommendation maps (Rxs) to their farmer-clients. This study proposes a method to empirically evaluate the efficacy of commercial Rxs. The method takes three steps: (1) it uses precision agriculture technology to conduct randomized on-farm precision experiments; (2) it estimates yield response functions for the Rx’s management zones using the data; and (3) it conducts economic analysis to test the hypothesis that implementing the Rx is an economically optimal strategy. The method is illustrated using data from a 2018 on-farm nitrogen and seed rate precision experiment on a 31-ha Ohio field, for which nitrogen and seed Rxs were created by the farmer’s professional consultant. The study demonstrates the promise of improving input management through on-farm precision experimentation and data analysis. Future research must conduct trials over multiple years to account for weather. A call is made for the development of public and private research infrastructure to lower the costs on-farm precision experimentation and data analysis.

Data-centric technology has not undergone widespread adoption in production agriculture but could address global needs for food security and farm profitability. Participants in the U.S. Department of Agriculture (USDA) National Institute for Food and Agriculture (NIFA) funded conference, “Identifying Obstacles to Applying Big Data in Agriculture,” held in Houston, TX, in August 2018, defined detailed scenarios in which on-farm decisions could benefit from the application of Big Data. The participants came from multiple academic fields, agricultural industries and government organizations and, in addition to defining the scenarios, they identified obstacles to implementing Big Data in these scenarios as well as potential solutions. This communication is a report on the conference and its outcomes. Two scenarios are included to represent the overall key findings in commonly identified obstacles and solutions: “In-season yield prediction for real-time decision-making”, and “Sow lameness.” Common obstacles identified at the conference included error in the data, inaccessibility of the data, unusability of the data, incompatibility of data generation and processing systems, the inconvenience of handling the data, the lack of a clear return on investment (ROI) and unclear ownership. Less common but valuable solutions to common obstacles are also noted.

This study was conducted to model the fraction of intercepted photosynthetically active radiation (fIPAR) in heterogeneous row-structured orchards, and to develop methodologies for accurate mapping of the instantaneous fIPAR at field scale using remote sensing imagery. The generation of high-resolution maps delineating the spatial variation of the radiation interception is critical for precision agriculture purposes such as adjusting management actions and harvesting in homogeneous within-field areas. Scaling-up and model inversion methods were investigated to estimate fIPAR using the 3D radiative transfer model, Forest Light Interaction Model (FLIGHT). The model was tested against airborne and field measurements of canopy reflectance and fIPAR acquired on two commercial peach and citrus orchards, where study plots showing a gradient in the canopy structure were selected. High-resolution airborne multi-spectral imagery was acquired at 10 nm bandwidth and 150 mm spatial resolution using a miniaturized multi-spectral camera on board an unmanned aerial vehicle (UAV). In addition, simulations of the land surface bidirectional reflectance were conducted to understand the relationships between canopy architecture and fIPAR. Input parameters used for the canopy model, such as the leaf and soil optical properties, canopy architecture, and sun geometry were studied in order to assess the effect of these inputs on canopy reflectance, vegetation indices and fIPAR. The 3D canopy model approach used to simulate the discontinuous row-tree canopies yielded spectral RMSE values below 0.03 (visible region) and below 0.05 (near-infrared) when compared against airborne canopy reflectance imagery acquired over the sites under study. The FLIGHT model assessment conducted for fIPAR estimation against field measurements yielded RMSE values below 0.08. The simulations conducted suggested the usefulness of these modeling methods in heterogeneous row-structured orchards, and the high sensitivity of the normalized difference vegetation index and fIPAR to background, row orientation, percentage cover and sun geometry. Mapping fIPAR from high-resolution airborne imagery through scaling-up and model inversion methods conducted with the 3D model yielded RMSE error values below 0.09 for the scaling-up approach, and below 0.10 for the model inversion conducted with a look-up table. The generation of intercepted radiation maps in row-structured tree orchards is demonstrated to be feasible using a miniaturized multi-spectral camera on board UAV platforms for precision agriculture purposes.

Coffee leaf rust (CLR) caused by the fungus Hemileia vastarix is a devastating disease in almost all coffee producing countries and remote sensing approaches have the potential to monitor the disease. This study evaluated the potential of Sentinel-2 band settings for discriminating CLR infection levels at leaf levels. Field spectra were resampled to the band settings of the Sentinel-2, and evaluated using the random forest (RF) and partial least squares discriminant analysis (PLS-DA) algorithms with and without variable optimization. Using all variables, Sentinel-2 Multispectral Imager (MSI)-derived vegetation indices achieved higher overall accuracy of 76.2% when compared to 69.8% obtained using raw spectral bands. Using the RF out-of-bag (OOB) scores, 4 spectral bands and 7 vegetation indices were identified as important variables in CLR discrimination. Using the PLS-DA Variable Importance in Projection (VIP) score, 3 Sentinel-2 spectral bands (B4, B6 and B5) and 5 vegetation indices were found to be important variables. Use of the identified variables improved the CLR discrimination accuracies to 79.4 and 82.5% for spectral bands and indices respectively when discriminated with the RF. Discrimination accuracy slightly increased through variable optimization for PLS-DA using spectral bands (63.5%) and vegetation indices (71.4%). Overall, this study showed the potential of the Sentinel 2 MSI band settings for CLR discrimination as part of crop condition assessment. Nevertheless further studies are required under field conditions.