Abstract
Knowledge of the factors influencing nutrient-limited subtropical maize yield and subsequent prediction is crucial for effective nutrient management, maximizing profitability, ensuring food security, and promoting environmental sustainability. We analyzed data from nutrient omission plot trials (NOPTs) conducted in 324 farmers' fields across ten agroecological zones (AEZs) in the Eastern Indo-Gangetic Plains (EIGP) of Bangladesh to explain maize yield variability and identify variables controlling nutrient-limited yields. An additive main effect and multiplicative interaction (AMMI) model was used to explain maize yield variability with nutrient addition. Interpretable machine learning (ML) algorithms in automatic machine learning (AutoML) frameworks were subsequently used to predict attainable yield relative nutrient-limited yield (RY) and to rank variables that control RY. The stack-ensemble model was identified as the best-performing model for predicting RYs of N, P, and Zn. In contrast, deep learning outperformed all base learners for predicting RY K . The best model's square errors (RMSEs) were 0.122, 0.105, 0.123, and 0.104 for RY N , RY P , RY K , and RY Zn , respectively. The permutation-based feature importance technique identified soil pH as the most critical variable controlling RY N and RY P . The RY K showed lower in the eastern longitudinal direction. Soil N and Zn were associated with RY Zn . The predicted median RY of N, P, K, and Zn, representing average soil fertility, was 0.51, 0.84, 0.87, and 0.97, accounting for 44, 54, 54, and 48% upland dry season crop area of Bangladesh, respectively. Efforts are needed to update databases cataloging variability in land type inundation classes, soil characteristics, and INS and combine them with farmers' crop management information to develop more precise nutrient guidelines for maize in the EIGP. • >70% of fields showed positive yield responses to N, P, K, and Zn across all AEZs. • Median nutrient limiting yields relative to balanced fertilization followed the order Zn > K > P > N. • The AMMI model is used to explain maize yield variability across Bangladesh. • The ensemble model was identified as the best-performing model for predicting RY. • The predicted RY accounting for 44–54% of the upland dry season crop area of Bangladesh.
Abstract from the paper’s OpenAlex record; open copy licensed cc-by-nc-nd. 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 16
Claim · model-written summary
A stack-ensemble model performed best for predicting relative yield for N, P, and Zn.
Zea maysOrganisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“The stack-ensemble model was identified as the best-performing model for predicting RYs of N, P, and Zn”
t0“maize yield”
Claim · model-written summary
Deep learning outperformed all base learners in predicting relative yield for K.
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“deep learning outperformed all base learners for predicting RY K”
t0“maize yield”
Claim · model-written summary
The best model had an RMSE of 0.122 for relative yield under N limitation.
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“The best model's square errors (RMSEs) were 0.122, 0.105, 0.123, and 0.104 for RY N , RY P , RY K , and RY Zn , respectively”
t0“maize yield”
Claim · model-written summary
The best model had an RMSE of 0.105 for relative yield under P limitation.
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“The best model's square errors (RMSEs) were 0.122, 0.105, 0.123, and 0.104 for RY N , RY P , RY K , and RY Zn , respectively”
t0“maize yield”
Claim · model-written summary
The best model had an RMSE of 0.123 for relative yield under K limitation.
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“The best model's square errors (RMSEs) were 0.122, 0.105, 0.123, and 0.104 for RY N , RY P , RY K , and RY Zn , respectively”
t0“maize yield”
Claim · model-written summary
The best model had an RMSE of 0.104 for relative yield under Zn limitation.
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“The best model's square errors (RMSEs) were 0.122, 0.105, 0.123, and 0.104 for RY N , RY P , RY K , and RY Zn , respectively”
t0“maize yield”
Claim · model-written summary
Soil pH was the most important variable controlling maize relative yields under N and P limitation.
Not reportedDirection
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“soil pH as the most critical variable controlling RY N and RY P”
t0“maize yield”
Claim · model-written summary
Maize relative yield under K limitation was lower toward the east.
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“The RY K showed lower in the eastern longitudinal direction”
t0“maize yield”
Claim · model-written summary
Soil N and Zn were associated with maize relative yield under Zn limitation.
Not reportedDirection
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“Soil N and Zn were associated with RY Zn”
t0“maize yield”
Claim · model-written summary
Predicted median relative yield under N limitation was 0.51, over 44% of upland dry-season cropland.
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“The predicted median RY of N, P, K, and Zn, representing average soil fertility, was 0.51, 0.84, 0.87, and 0.97, accounting for 44, 54, 54, and 48% upland dry season crop area of Bangladesh, respectively”
t0“maize yield”
Claim · model-written summary
Predicted median relative yield under P limitation was 0.84, over 54% of upland dry-season cropland.
- Outcome
- RY of N, Pe0
- Value
- 0.84e0
- Qualifier
- upland dry season crop area of Bangladeshe0
- Setting
- Bangladeshe0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“The predicted median RY of N, P, K, and Zn, representing average soil fertility, was 0.51, 0.84, 0.87, and 0.97, accounting for 44, 54, 54, and 48% upland dry season crop area of Bangladesh, respectively”
t0“maize yield”
Claim · model-written summary
Predicted median relative yield under K limitation was 0.87, over 54% of upland dry-season cropland.
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“The predicted median RY of N, P, K, and Zn, representing average soil fertility, was 0.51, 0.84, 0.87, and 0.97, accounting for 44, 54, 54, and 48% upland dry season crop area of Bangladesh, respectively”
t0“maize yield”
Claim · model-written summary
Predicted median relative yield under Zn limitation was 0.97, over 48% of upland dry-season cropland.
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“The predicted median RY of N, P, K, and Zn, representing average soil fertility, was 0.51, 0.84, 0.87, and 0.97, accounting for 44, 54, 54, and 48% upland dry season crop area of Bangladesh, respectively”
t0“maize yield”
Claim · model-written summary
More than 70% of fields showed a positive maize-yield response to N, P, K, and Zn.
Not reportedComparator
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“ >70% of fields showed positive yield responses to N, P, K, and Zn across all AEZs”
t0“maize yield”
Claim · model-written summary
Median nutrient-limited relative yields ranked Zn above K above P above N.
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“Median nutrient limiting yields relative to balanced fertilization followed the order Zn > K > P > N”
t0“maize yield”
Claim · model-written summary
The authors call for updated land, soil, and indigenous nutrient supply databases combined with management data to refine maize nutrient guidelines.
- Outcome
- develop more precise nutrient guidelines for maize in the EIGPe0
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“Efforts are needed to update databases cataloging variability in land type inundation classes, soil characteristics, and INS and combine them with farmers' crop management information to develop more precise nutrient guidelines for maize in the EIGP”
The same questions in other papers
zinc atom · yield · Zea mays (maize)Results differ
390 papers in the same direction as this paper · 47 in the opposite direction · 3 reporting no difference from an alternative
- negative
- negative
- no effect
437 papers more in this group
See all 441 papers →potassium · yield · Zea mays (maize)Positive and negative results
290 papers in the same direction as this paper · 28 in the opposite direction · 1 reporting no difference from an alternative
- no effect
- negative
- no effect
316 papers more in this group
See all 320 papers →phosphorus · yield · Zea mays (maize)Positive and negative results
548 papers in the same direction as this paper · 69 in the opposite direction · 9 reporting no difference from an alternative
- no effect
- no effect
- negative
623 papers more in this group
See all 627 papers →nitrogen · yield · Zea mays (maize)Positive and negative results
1,905 papers in the same direction as this paper · 234 in the opposite direction · 45 reporting no difference from an alternative
- negative
- no effect
- negative
2,181 papers more in this group
See all 2,185 papers →Claims are grouped by organism, treatment and what was measured or targeted. A paper counts as positive when any of its claims in the group is positive.