Development of a new cold hardiness prediction model for grapevine using phased integration of acclimation and deacclimation responses

Kovaleski, Al P.; North, Michael G.; Martinson, Timothy E.; Londo, Jason P.

doi:10.1016/j.agrformet.2023.109324

Cited by 7 publications

(17 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…3). This reflects temperature conditions in each treatment (lower temperatures experienced = greater cold hardiness) and was expected based on models used for prediction of cold hardiness of grapevines (Ferguson et al, 2011, 2014; North et al, 2022; Kovaleski et al, 2023; Jones et al, 2023).…”

Section: Discussionmentioning

confidence: 80%

“…However, artificial chilling treatments are typically not as cold as most temperate environments, even when considering studies that have included freezing temperatures (Mahmood et al, 2000;Rose and Cameron, 2009;Guak and Neilsen, 2013;Cragin et al, 2017;Baumgarten et al, 2021). Based on cold hardiness models, warmer temperatures lead to smaller rates of cold hardiness gain (Ferguson et al, 2011(Ferguson et al, , 2014North et al, 2022;Kovaleski et al, 2023;Jones et al, 2023), but can also not elicit a genotypes' full potential in terms of cold hardiness (Kovaleski et al, 2023;Jones et al, 2023). Thus, different artificial chilling treatments likely promote uneven levels of cold acclimation.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Cold hardiness-informed budbreak reveals role of freezing temperatures and daily fluctuation in chill accumulation model

North,

Workmaster,

Atucha

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

Fundamental questions in bud dormancy remain, including what temperatures fulfill dormancy requirements (i.e., chill accumulation). Recent studies demonstrate freezing temperatures promote chill accumulation and cold hardiness influences time to budbreak — the phenotype used for dormancy evaluations. Here we evaluated bud cold hardiness (CH) and budbreak responses of grapevines (Vitis hybrids) throughout chill accumulation under three treatments: constant (5 °C), fluctuating (-3.5 to 6.5 °C daily), and field conditions (Madison, WI, USA). Chill treatments experiencing lower temperatures promoted greater gains in cold hardiness (CHfield>CHfluctuating>CHconstant). All treatments decreased observed time to budbreak with increased chill accumulation. However, perceived treatment effectiveness changed when time to budbreak was adjusted to remove cold acclimation effects. Among three classic chill models (North Carolina, Utah, and Dynamic), none were able to correctly describe adjusted time to budbreak responses to chill accumulation. Thus, a new model is proposed that expands the range of chill accumulation temperatures to include freezing temperatures and enhances chill accumulation under fluctuating temperature conditions. Most importantly, our analysis demonstrates adjustments for uneven acclimation change the perceived effectiveness of chill treatments. Therefore, future work in bud dormancy would benefit from simultaneously evaluating cold hardiness.

show abstract

Section: Discussionmentioning

confidence: 80%

Section: Introductionmentioning

confidence: 99%

Cold hardiness-informed budbreak reveals role of freezing temperatures and daily fluctuation in chill accumulation model

North,

Workmaster,

Atucha

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…S1 ). These results demonstrated an improvement over predictions produced by the NYUS.1, WAUS.2, and the RNN model during the testing on unseen data [ 14 , 34 ], indicating that the alpha model passed the method suitability test and the site-transferability test. As the internal and external testing data are from different sites, this result also suggests that the AutoGluon method did not overfit the training data and exhibited high site-transferability.…”

Section: Discussionmentioning

confidence: 99%

“…Chilling accumulation has been determined for its importance in the physiology of grapevine in the dormant season [ 4 , 13 , 46 , 47 ]. Different chilling models also help quantify the dynamics of cold acclimation and deacclimation in the NYUS.1 and WAUS.2 models [ 11 , 14 ], and therefore errors from chilling models can hinder predictions. The Auto-ML model utilized individual chilling models in a different manner than the previous prediction models ( Fig.…”

Section: Discussionmentioning

confidence: 99%

“…First, the parameters used in the WAUS.2 and NYUS.1 models were optimized using local freezing tolerance data in Washington and New York, respectively, which may contribute to overfitting under local climate conditions and underperformance under other climate conditions. For example, the underperformance of WAUS.2 has been observed in New York, Wisconsin, and in subtropical viticultural regions in Chile [ 12 , 14 , 16 ]. Second, since these models are the mathematical realizations of the biological understanding of dormancy and freezing tolerance, the parameters in the models such as the dynamics of cold acclimation and deacclimation need to be updated regularly according to the advances in dormancy and freezing tolerance biology.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

NYUS.2: an automated machine learning prediction model for the large-scale real-time simulation of grapevine freezing tolerance in North America

Wang,

Moghe,

Kovaleski

et al. 2023

Horticulture Research

View full text Add to dashboard Cite

Accurate and real-time monitoring of grapevine freezing tolerance is crucial for the sustainability of the grape industry in cool climate viticultural regions. However, on-site data is limited due to the complexity of measurement. Current prediction models underperform under diverse climate conditions, which limits the large-scale deployment of these methods. We combined grapevine freezing tolerance data from multiple regions in North America and generated a predictive model based on hourly temperature-derived features and cultivar features using AutoGluon, an automated machine learning engine. Feature importance was quantified by AutoGluon and SHAP value. The final model was evaluated and compared with previous models for its performance under different climate conditions. The final model achieved an overall 1.36 °C root-mean-square error during model testing and outperformed two previous models using three test cultivars at all testing regions. Two feature importance quantification methods identified five shared essential features. Detailed analysis of the features indicates that the model has adequately extracted some biological mechanisms during training. The final model, named NYUS.2, was deployed along with two previous models as an R shiny-based application in the 2022-2023 dormancy season, enabling large-scale and real-time simulation of grapevine freezing tolerance in North America for the first time.

show abstract

Cold hardiness-informed budbreak reveals role of freezing temperatures and daily fluctuation in chill accumulation model

North,

Workmaster,

Atucha

et al. 2024

Journal of Experimental Botany

View full text Add to dashboard Cite

Fundamental questions in bud dormancy remain, including what temperatures fulfill dormancy requirements (i.e., chill accumulation). Recent studies demonstrate freezing temperatures promote chill accumulation and cold hardiness influences time to budbreak – the phenotype used for dormancy evaluations. Here we evaluated bud cold hardiness (CH) and budbreak responses of grapevines (Vitis hybrids) throughout chill accumulation under three treatments: constant (5 °C), fluctuating (-3.5 to 6.5 °C daily), and field conditions (Madison, WI, USA). Chill treatments experiencing lower temperatures promoted greater gains in cold hardiness (CHfield>CHfluctuating>CHconstant). All treatments decreased observed time to budbreak with increased chill accumulation. However, perceived treatment effectiveness changed when time to budbreak was adjusted to remove cold acclimation effects. Among three classic chill models (North Carolina, Utah, and Dynamic), none were able to correctly describe adjusted time to budbreak responses to chill accumulation. Thus, a new model is proposed that expands the range of chill accumulation temperatures to include freezing temperatures and enhances chill accumulation under fluctuating temperature conditions. Most importantly, our analysis demonstrates adjustments for uneven acclimation change the perceived effectiveness of chill treatments. Therefore, future work in bud dormancy would benefit from simultaneously evaluating cold hardiness.

show abstract

Development of a new cold hardiness prediction model for grapevine using phased integration of acclimation and deacclimation responses

Cited by 7 publications

References 36 publications

Cold hardiness-informed budbreak reveals role of freezing temperatures and daily fluctuation in chill accumulation model

Cold hardiness-informed budbreak reveals role of freezing temperatures and daily fluctuation in chill accumulation model

NYUS.2: an automated machine learning prediction model for the large-scale real-time simulation of grapevine freezing tolerance in North America

Cold hardiness-informed budbreak reveals role of freezing temperatures and daily fluctuation in chill accumulation model

Contact Info

Product

Resources

About