Search

Irrigation scheduling based on the crop water stress index (CWSI) and temperature-time threshold (TTT) methods is promising for semi-arid and arid climates. The objective of this study was to investigate if CWSI and TTT methods could be used as irrigation signaling tools for a humid environment in the southeastern U.S. Corn canopy temperature data were collected in Alabama in 2018 and 2019 using infrared leaf temperature sensors on a fully irrigated treatment and on two limited irrigation treatments. A set of three soil water sensors installed at 0.15, 0.3, and 0.6 m soil depth were used to prescribe irrigation time and amount. CWSI was sensitive to precipitation, irrigation, and plant water uptake. No statistical differences in CWSI or yield among the three irrigation levels were found in 2018 when precipitation was well distributed during the season. In contrast, during 2019 both CWSI and yield differed significantly among the three irrigation treatments. Precipitation events in 2019 were sparse compared to 2018; therefore, irrigation promoted greater differences in water availability between treatments. Inconsistencies observed in potential irrigation signaling using the TTT method with or without the inclusion of a limiting relative humidity algorithm indicate that the TTT method may not be a reliable irrigation signaling tool for humid environments. [ABSTRACT FROM AUTHOR]

The implementation of adequate irrigation strategies could be done through real-time monitoring of soil water status at several soil depths; however, this could also represent a complex nonlinear problem due to the plant-soil-weather relationships. In this study, two recurrent neural network (RNN) models were evaluated to estimate irrigation prescriptions. Data for this study were collected from an on-farm corn irrigation study conducted between 2017 and 2019 in Samson, Alabama. The study used hourly data of weather and soil matric potential (SMP) monitored at three soil depths from 13 sensor probes installed on a loamy fine sand soil and a sandy clay loam soil. Two neural network methods, i.e., a nonlinear autoregressive with exogenous (NARX) input system and long short-term memory (LSTM), were trained, validated, and tested with a maximum dataset of 20,052 records and a maximum of eight categorical attributes to estimate one-step irrigation prescriptions. The performance of both methods was evaluated by varying the model development parameters (neurons or blocks, dropout, and epochs) and determining their impact on the final model prediction. Results showed that both RNN models demonstrated good capability in the prediction of irrigation prescriptions for the soil types studied, with a coefficient of determination (R2) > 0.94 and root mean square error (RMSE) < 1.2 mm. The results of this study indicate that after training the RNNs using the dataset collected in the field, models using only SMP sensors at three soil depths obtained the best performance, followed by models that used only data of solar radiation, temperature, and relative humidity in the prediction of irrigation prescriptions. For future applicability, the RNN models can be extended using datasets from other places for training, which would allow the adoption of a unique data-driven soil moisture model for irrigation scheduling useful in a wide range of soil types. [ABSTRACT FROM AUTHOR]

Management zones (MZs) for southern root-knot nematode (RKN) from the integration of terrain (TR) and edaphic (ED) field features might facilitate variable rate nematicide applications. This study was conducted on 11 coastal plain fields in the USA. The relationships between RKN populations and five soil ED and TR attributes (apparent soil electrical conductivity [shallow (ECa-s) and deep (ECa-d)], elevation (EL), slope (SL), and changes in bare soil reflectance) were analyzed using canonical correlation. Using two ED and TR data sets, canonical predictors were used for zone delineation. Although the results showed that the zones with RKN population above the RKN field average were associated with the lowest values of ECa-s, ECa-d, normalized difference vegetation index (NDVI), and SL with respect to field average values, zone segregation was enough using ECa-s and ECa-d data. The results suggest the potential for using soil properties to identify RKN risk zones. [ABSTRACT FROM AUTHOR]

Cotton (Gossypium hirsutum L.) yield losses by southern root-knot nematode (RKN; Meloidogyne incognita (Kofoid & White) Chitwood) are usually assessed after significant damage has been caused. However, estimation of potential yield reduction before planting is possible by using crop simulation. The main goal of this study was to adapt the Cropping System Model (CSM)-CROPGRO-Cotton for simulating growth and yield of cotton plants infected with RKN. Two hypotheses were evaluated to simulate RKN damage: (1) RKN acting as a sink for soluble assimilate, and (2) RKN inducing a reduction of root length per root mass and root density. The model was calibrated and adapted using data collected in an experiment that was conducted in 2007 and was part of a long-term crop rotation study. The experiment had a split-plot design, replicated six times, with drought stress levels assigned to the main plots and fumigation levels assigned to the subplots. The model was evaluated with seed cotton weight data collected in an experiment that was conducted in 2001 and was part of the same long-term crop rotation experiment. The fumigation treatments created various levels of RKN population densities. The model was adapted by coupling the RKN population to the removal of daily assimilates and decreasing root length per unit mass. The assimilate consumption rate was obtained after minimizing the error between simulated and observed biomass and yield components for the limited drought stress, non-fumigated treatment. Different values of root length per unit root weight (RFAC1) were used to account for early symptoms of RKN damage on leaf area index (LAI) and vegetative biomass under the non-fumigated, drought stress conditions. After model adaptation, the simulations indicated that lAI, total biomass, boll weight, and seed cotton decreased with elevated RKN population. The impact of RKN was more pronounced under severe drought stress. The lowest RMSE of LAI simulations occurred for the non-fumigated treatments under medium and severe drought stress (0. 71 and 0.65 m² m-2, respectively). Biomass was simulated with a prediction error within a range of 6% to 18. 4% and seed cotton within a range of-11.2% to 2. 7%. Seed cotton weight losses associated with RKN infection increased with the level of drought stress (9%, 20%, and 18% for the low, medium, and severe drought stress). Model evaluation showed that seed cotton weight was slightly more overpredicted for the fumigated than for the non-fumigated treatments, with prediction errors of 28.2%, 15.8%, and 2.0% for the low, medium, and severe drought stress, respectively. Similar to the calibration of the model, the yield losses increased with the combination of RKN and drought stress (20% and 29% for the low and severe drought stress). The results showed the potential for using the CSM-CROPGRO-Cotton model to account for RKN damage as well as to simulate yield reduction. However, further model evaluation might be needed to evaluate the values of assimilate consumption and root length per unit weight for different environmental conditions and management practices. [ABSTRACT FROM AUTHOR]

Identifying the spatial variability and risk areas for southern root-knot nematode [Meloidogyne incognita (Kofoid & White) Chitwood] (RKN) is key for site-specific management (SSM) of cotton (Gossypium hirsutum L.) fields. The objectives of this study were to: (i) determine the soil properties that influence RKN occurrence at different scales; and (ii) delineate risk areas of RKN by indicator kriging. The study site was a cotton field located in the southeastern coastal plain region of the USA. Nested semivariograms indicated that RKN samples, collected from a 50×50 m grid, exhibited a local and regional scale of variation describing small and large clusters of RKN population density. Factorial kriging decomposed RKN and soil properties variability into different spatial components. Scale dependent correlations between RKN data showed that the areas with high RKN population remained stable though the growing season. RKN data were strongly correlated with slope (SL) at local scale and with apparent soil electrical conductivity deep (EC(a-d)) at both local and regional scales, which illustrate the potential of these soil physical properties as surrogate data for RKN population. The correlation between RKN data and soil chemical properties was soil texture mediated. Indicator kriging (IK) maps developed using either RKN, the relation between RKN and soil electrical conductivity or a combination of both, depicted the probability for RKN population to exceed the threshold of 100 second stage juveniles/100 cm(3) of soil. Incorporating EC(a-d) as soft data improved predictions favoring the reduction of the number of RKN observations required to map areas at risk for high RKN population.