Journal of Agronomy and Field Innovations

Background: Glycine max (L.) Merr. (soybean) is one of the most important oilseeds crops worldwide, and it is experiencing unprecedented challenges due to climate change, resulting in drought and water stress. Conventional crop monitoring approaches are not sufficient in terms of their spatial resolution and timing for the precision management of irrigation.
Objective: To estimate the effectiveness of drone-based multispectral imaging for determining the water stress in soybeans at an early stage of development ahead of the visual symptoms appearance and thereby applying the timely irrigation measures.
Methods: The two-season field experiment applied the randomized complete block method and separated two treatments: well-watered and stressed plants. The multispectral images were taken using the DJI Matrice 300 RTK when the height was 50 m, and MicaSense RedEdge-MX camera was used at vegetative, flowering, pod development, and seed-filling stages. The vegetation indices (NDVI, NDRE, GNDVI, SAVI, MSAVI) were calculated using Pix4Dmapper software and their correlations to the ground-truth measurements such as leaf water potential, relative water content, chlorophyll content, and stomatal conductance were analyzed.
Result: In terms of the results, it was noted that the values of the multispectral vegetation indices showed significant variations dependent on time (i.e. P value was below 0.05). At the same time, NDRE and MSAVI were found to be the most effective in identifying the early stages of water stress (with a coefficient of determination of 0.89). Plants experiencing water stress could be distinguished from those that were well watered 7–10 days before any visible signs took place. The seed yield in water-stressed plots was lower than in the well-watered ones by 34% (i.e. 1.85 Mg/ha and 2.81 Mg/ha respectively), and there were strong connections between the vegetation indices and seed yield (r = 0.92, P < 0.001).
Conclusion: As for the conclusion, it can be stated that the drone-based multispectral imaging provides the efficient, rapid, and not expensive method of water stress detection in soybean and increases the efficiency of water consumption by 18% if used under perfect conditions.