Abstract
Timely and accurate prediction of crop traits is critical for precision breeding and regional agricultural production. Previous studies have primarily focused on single crop yield traits, neglecting other crop traits and variety-specific analyses. To address this issue, we employed a Meta-Hybrid Regression Ensemble (MHRE) approach by using multiple machine learning (ML) approaches as base learners, integrating regional multi-year, multi-variety crop field trials with satellite remote sensing indices, meteorological and phenological data to predict major crop traits. Results demonstrated MHRE’s optimal performance for rice and cotton, significantly outperforming individual models (RF, XGBoost, CatBoost, and LightGBM). Specifically, for rice crop, MHRE achieved highest accuracy for yield trait (R2 = 0.78, RMSE = 0.59 t ha−1) compared to the best individual model (XGBoost: R2 = 0.76, RMSE = 0.61 t ha−1); traits like effective spike also showed strong predictability (R2 = 0.64, RMSE = 27.81 10,000·spike ha−1). Similarly, for cotton, MHRE substantially improved yield trait prediction (R2 = 0.82, RMSE = 0.33 t ha−1) compared to the best individual model (RF: R2 = 0.77, RMSE = 0.36 t ha−1); bolls per plant accuracy was highest (R2 = 0.93, RMSE = 2.27 bolls plant−1). Moreover, rigorous validation confirmed that crop-specific MHRE models are robust across five rice and three cotton varietal groups and are applicable across six distinct regions in China. Furthermore, we applied the SHAP (SHapley Additive exPlanations) method to analyze the growth stages and key environmental factors affecting major traits. Our study illustrates a practical framework for regional-scale crop traits prediction by fusing multi-source data and ensemble machine learning, offering new insights for precision agriculture and crop management.
Abstract from the paper’s OpenAlex record; open copy licensed cc-by. Read at the source ↗
The text the model read, with the source record’s inline markup (italics, sub- and superscripts) rendered and its character entities decoded; quote offsets refer to the raw text.
Claims 8
Claim · model-written summary
The MHRE ensemble performed best for rice and cotton and significantly outperformed the individual models RF, XGBoost, CatBoost and LightGBM.
- Taxon
- rice and cotton
Gossypium hirsutum
Oryza sativae1 - Outcome
- optimal performancee1
- Direction
- outperforminge1
- Comparator
- individual modelse1
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e1“Results demonstrated MHRE’s optimal performance for rice and cotton, significantly outperforming individual models (RF, XGBoost, CatBoost, and LightGBM)”
Claim · model-written summary
For rice yield, MHRE reached the highest accuracy (R2 = 0.78, RMSE = 0.59 t ha−1) versus the best individual model XGBoost (R2 = 0.76, RMSE = 0.61 t ha−1).
- Taxon
- riceOryza sativae1
- Outcome
- accuracy for yield traite1
- Direction
- higheste1
- Value
- R2 = 0.78, RMSE = 0.59 t ha−1e1
- Qualifier
- XGBoost: R2 = 0.76, RMSE = 0.61 t ha−1e1
- Comparator
- best individual modele1
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e1“for rice crop, MHRE achieved highest accuracy for yield trait (R2 = 0.78, RMSE = 0.59 t ha−1) compared to the best individual model (XGBoost: R2 = 0.76, RMSE = 0.61 t ha−1)”
Claim · model-written summary
Effective spike number in rice also showed strong predictability (R2 = 0.64, RMSE = 27.81 10,000·spike ha−1).
- Taxon
- RiceOryza sativat0
- Outcome
- effective spikee1
- Value
- R2 = 0.64, RMSE = 27.81 10,000·spike ha−1e1
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e1“traits like effective spike also showed strong predictability (R2 = 0.64, RMSE = 27.81 10,000·spike ha−1)”
t0“Rice and Cotton”
Claim · model-written summary
For cotton yield, MHRE substantially improved prediction (R2 = 0.82, RMSE = 0.33 t ha−1) over the best individual model RF (R2 = 0.77, RMSE = 0.36 t ha−1).
- Taxon
- cottonGossypium hirsutume1
- Outcome
- yield trait predictione1
- Direction
- substantially improvede1
- Value
- R2 = 0.82, RMSE = 0.33 t ha−1e1
- Qualifier
- RF: R2 = 0.77, RMSE = 0.36 t ha−1e1
- Comparator
- best individual modele1
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e1“for cotton, MHRE substantially improved yield trait prediction (R2 = 0.82, RMSE = 0.33 t ha−1) compared to the best individual model (RF: R2 = 0.77, RMSE = 0.36 t ha−1)”
Claim · model-written summary
For cotton, accuracy for bolls per plant was the highest of the traits (R2 = 0.93, RMSE = 2.27 bolls plant−1).
- Taxon
- CottonGossypium hirsutumt0
- Outcome
- bolls per plant accuracye1
- Direction
- higheste1
- Value
- R2 = 0.93, RMSE = 2.27 bolls plant−1e1
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e1“bolls per plant accuracy was highest (R2 = 0.93, RMSE = 2.27 bolls plant−1)”
t0“Rice and Cotton”
Claim · model-written summary
The crop-specific MHRE model for rice was robust across five rice varietal groups.
- Taxon
- riceOryza sativae1
- Outcome
- robuste1
- Value
- five ricee1
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e1“robust across five rice and three cotton varietal groups”
Claim · model-written summary
The crop-specific MHRE model for cotton was robust across three cotton varietal groups.
- Taxon
- cottonGossypium hirsutume1
- Outcome
- robuste1
- Value
- three cottone1
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e1“robust across five rice and three cotton varietal groups”
Claim · model-written summary
The crop-specific MHRE models were applicable across six distinct regions in China.
Verbatim
e1“applicable across six distinct regions in China”