Soybean prediction using computationally efficient Bayesian spatial regression models and satellite imagery DOI

Richard J. Fischer,

Hossein Moradi Rekabdarkolaee, Deepak R. Joshi

et al.

Agronomy Journal, Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 3, 2024

Abstract Preharvest yield estimates can be used for harvest planning, marketing, and prescribing in‐season fertilizer pesticide applications. One approach that is being widely tested the use of machine learning (ML) or artificial intelligence (AI) algorithms to estimate yields. However, one barrier adoption this ML/AI behave as a black block. An alternative create an algorithm using Bayesian statistics. In statistics, prior information help algorithm. based on statistics are not often computationally efficient. The objective current study was compare accuracy computational efficiency four models different assumptions reduce execution time. paper, multiple linear regression (BLR), spatial, skewed spatial regression, nearest neighbor Gaussian process (NNGP) were compared with ML non‐Bayesian random forest model. analysis, soybean ( Glycine max ) yields response variable y ), spaced‐based blue, green, red, near‐infrared reflectance measured PlanetScope satellite predictor x ). Among tested, (NNGP; R 2 ‐testing = 0.485) model, which captures short‐range correlation, outperformed (BLR; 0.02), (SRM; 0.087), (sSRM; 0.236) models. associated improved increase in run time from 534 s BLR model 2047 NNGP These data show relatively accurate within‐field obtained without sacrificing coefficients have biological meaning. all had lower values higher times than

Language: Английский

Plant organ rather than cover crop species determines residue incorporation into SOC pools DOI Creative Commons

Tine Engedal,

Veronika Hansen, Jim Rasmussen

et al.

Soil Biology and Biochemistry, Journal Year: 2024, Volume and Issue: 200, P. 109616 - 109616

Published: Oct. 5, 2024

Language: Английский

Citations

2

Using Beerkan Procedure to Estimate Hydraulic Soil Properties under Long Term Agroecosystems Experiments DOI Creative Commons
Lorenzo Vergni, Grazia Tosi,

Jennifer Bertuzzi

et al.

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(9), P. 3817 - 3817

Published: April 30, 2024

The BEST (Beerkan Estimation of Soil Transfer parameters) method was used to compare the hydraulic properties soils in two Long-term Agroecosystem Experiments (LTAEs) located at FIELDLAB experimental site University Perugia (central Italy). LTAE “NewSmoca” consists a biennial maize-durum wheat crop rotation under integrated low-input cropping systems with (i) inversion soil tillage (INT) or (ii) no-tillage (INT+) and (iii) an organic system (ORG). ORG INT+ involve use autumn-sown cover crops (before maize cycle). Pure stand durum grown INT INT+, while faba bean–wheat temporary intercropping implemented ORG. “Crop Rotation” different rotations residue management, continuous soft winter bean. Each is combined modes management: removal burial. For despite high-bulk density (>1.50 g/cm3), we found that conductivity, sorptivity available water are comparable those INT, probably due more structured efficient micropore system. show highest content values, recent spring occurring inter-row bean incorporation into soil. Rotation, burial seems influence capacity-based indicators positively. However, differences treatment minor, this could be tillage, which limits progressive accumulation matter.

Language: Английский

Citations

1

Understanding farmers’ adoption of diversified crop rotations in South Dakota, USA: an examination of the roles of sense of place and social responsibility DOI
Edem Avemegah, Elizabeth A. Bennett, Jessica D. Ulrich‐Schad

et al.

Agroecology and Sustainable Food Systems, Journal Year: 2024, Volume and Issue: 48(7), P. 934 - 960

Published: May 7, 2024

Diversified crop rotations can be a win-win solution for farmers and society given increased agronomic yield improved ecosystem services. However, the adoption of sustainable production practices must widespread accelerated to create resilient agroecosystems that remain productive as climate changes. In this paper, we use modified measures sense place (SOP) social responsibility (SR) investigate factors influence diversified (DCR) among producers in South Dakota (SD). Data was collected from 34 SD counties east Missouri River. Through application exploratory factor analysis (EFA), identified 3 constructs SOP on working landscapes 1) attachment identity, 2) networks, 3) physical dependence. Using binary logistic regression, positive association between DCR identity found. This suggests have an emotional bond their land plays role usage DCR. Our results suggest measuring some dimensions landscape context is important, but needs more refinement, specifically economic dependence, items it did not emerge EFA.

Language: Английский

Citations

1

Cover crops and deep‐soil C accumulation: What does research show after 10 years? DOI Creative Commons
Humberto Blanco‐Canqui,

Paul J. Jasa,

Richard B. Ferguson

et al.

Soil Science Society of America Journal, Journal Year: 2024, Volume and Issue: 88(6), P. 2167 - 2180

Published: Aug. 26, 2024

Abstract The extent to which cover crops (CCs) accumulate soil organic carbon (SOC) in the entire profile is still unclear. We measured SOC, permanganate oxidizable C (POX‐C), and particulate matter (POM) concentrations down 60‐cm depth early [2–3 week before corn ( Zea mays L.) planting]‐ late‐terminated (at planting) winter rye Secale cereale CCs rainfed irrigated no‐till continuous systems U.S. Corn Belt after 10 years. increased SOC stock POX‐C, POM but only system upper 5‐cm depth. Late‐terminated CC concentration by 4.710 ± 3.501 g kg −1 accumulated at 0.207 0.145 Mg ha year . It POX‐C concentrations, on average, 1.194 times. likely producing more biomass (2.247 0.370 ) than (0.949 0.338 ). At least 2 of may be needed increase SOC. Because often produce <1 when typically planted late terminated early, extending growing window terminating or crop planting (planting green) boost accumulation, although high‐C soils Mollisols, such as our study (>22 ), limit gains. submit would sequester low‐C, eroded, low‐fertility soils. Overall, minimally alter

Language: Английский

Citations

1

Soybean prediction using computationally efficient Bayesian spatial regression models and satellite imagery DOI

Richard J. Fischer,

Hossein Moradi Rekabdarkolaee, Deepak R. Joshi

et al.

Agronomy Journal, Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 3, 2024

Abstract Preharvest yield estimates can be used for harvest planning, marketing, and prescribing in‐season fertilizer pesticide applications. One approach that is being widely tested the use of machine learning (ML) or artificial intelligence (AI) algorithms to estimate yields. However, one barrier adoption this ML/AI behave as a black block. An alternative create an algorithm using Bayesian statistics. In statistics, prior information help algorithm. based on statistics are not often computationally efficient. The objective current study was compare accuracy computational efficiency four models different assumptions reduce execution time. paper, multiple linear regression (BLR), spatial, skewed spatial regression, nearest neighbor Gaussian process (NNGP) were compared with ML non‐Bayesian random forest model. analysis, soybean ( Glycine max ) yields response variable y ), spaced‐based blue, green, red, near‐infrared reflectance measured PlanetScope satellite predictor x ). Among tested, (NNGP; R 2 ‐testing = 0.485) model, which captures short‐range correlation, outperformed (BLR; 0.02), (SRM; 0.087), (sSRM; 0.236) models. associated improved increase in run time from 534 s BLR model 2047 NNGP These data show relatively accurate within‐field obtained without sacrificing coefficients have biological meaning. all had lower values higher times than

Language: Английский

Citations

1