ObjectivePrecise monitoring of cotton leaves’ nitrogen content is important for increasing yield and reducing fertilizer application. Spectra and images are used to monitor crop nitrogen information. However, the information expressed using nitrogen monitoring based on a single data source is limited and cannot consider the expression of various phenotypic and physiological parameters simultaneously, which can affect the accuracy of inversion. Introducing a multi-source data-fusion mechanism can improve the accuracy and stability of cotton nitrogen content monitoring from the perspective of information complementarity.MethodsFive nitrogen treatments were applied to the test crop, Xinluzao No. 53 cotton, grown indoors. Cotton leaf hyperspectral, chlorophyll fluorescence, and digital image data were collected and screened. A multilevel data-fusion model combining multiple machine learning and stacking integration learning was built from three dimensions: feature-level fusion, decision-level fusion, and hybrid fusion.ResultsThe determination coefficients (R2) of the feature-level fusion, decision-level fusion, and hybrid-fusion models were 0.752, 0.771, and 0.848, and the root-mean-square errors (RMSE) were 3.806, 3.558, and 2.898, respectively. Compared with the nitrogen estimation models of the three single data sources, R2 increased by 5.0%, 6.8%, and 14.6%, and the RMSE decreased by 3.2%, 9.5%, and 26.3%, respectively.ConclusionThe multilevel fusion model can improve accuracy to varying degrees, and the accuracy and stability were highest with the hybrid-fusion model; these results provide theoretical and technical support for optimizing an accurate method of monitoring cotton leaf nitrogen content.
The early and accurate monitoring of crop yield is important for field management, storage needs, and cash flow budgeting. Traditional cotton yield measurement methods are time-consuming, labor-intensive, and subjective. Chlorophyll fluorescence signals originate from within the plant and have the advantages of being fast and non-destructive, and the relevant parameters can reflect the intrinsic physiological characteristics of the plant. Therefore, in this study, the top four functional leaves of cotton plants at the beginning of the flocculation stage were used to investigate the pattern of the response of chlorophyll fluorescence parameters (e.g., F0, Fm, Fv/F0, and Fv/Fm) to nitrogen, and the cumulative fluorescence parameters were constructed by combining them with the leaf area index to clarify the correlation between chlorophyll fluorescence parameters and cotton yield. Support vector machine regression (SVM), an artificial neural network (BP), and an XGBoost regression tree were used to establish a cotton yield prediction model. Chlorophyll fluorescence parameters showed the same performance as photosynthetic parameters, which decreased as leaf position decreased. It showed a trend of increasing and then decreasing with increasing N application level, reaching the maximum value at 240 kg·hm-2 of N application. The correlation between fluorescence parameters and yield in the first, second, and third leaves was significantly higher than that in the fourth leaf, and the correlation between fluorescence accumulation and yield in each leaf was significantly higher than that of the fluorescence parameters, with the best performance of Fv/Fm accumulation found in the second leaf. The correlation between Fv/Fm accumulation and yield in the top three leaves combined was significantly higher than that in the top four leaves. The correlation coefficient between Fv/Fm accumulation and yield was the highest, indicating the feasibility of applying chlorophyll fluorescence to estimate yield. Based on the machine learning algorithm used to construct a cotton yield prediction model, the estimation models of Fv/F0 accumulation and yield of the top two leaves combined as well as top three leaves combined were superior. The estimation model coefficient of determination of the top two leaves combined in the BP algorithm was the highest. In general, the Fv/F0 accumulation of the top two leaves combined could more reliably predict cotton yield, which could provide technical support for cotton growth monitoring and precision management.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.