Developing a simple and efficient modeling solution for predicting key phenological stages of table grapes in a non-traditional viticulture zone in south Asia

Rizwan Rafique; Touqeer Ahmad; Muhammad Azam Khan; Mukhtar Ahmed; Gerrit Hoogenboom

doi:10.1007/s00484-024-02686-6

Developing a simple and efficient modeling solution for predicting key phenological stages of table grapes in a non-traditional viticulture zone in south Asia

Int J Biometeorol. 2024 May 9. doi: 10.1007/s00484-024-02686-6. Online ahead of print.

Authors

Rizwan Rafique^{1

2}, Touqeer Ahmad³, Muhammad Azam Khan², Mukhtar Ahmed⁴, Gerrit Hoogenboom^{1

5}

Affiliations

¹ Department of Agricultural and Biological Engineering, University of Florida, Gainesville, FL, USA.
² Horticulture Department, PMAS-Arid Agriculture University Rawalpindi, Rawalpindi, Pakistan.
³ Horticulture Department, PMAS-Arid Agriculture University Rawalpindi, Rawalpindi, Pakistan. touqeer@uaar.edu.pk.
⁴ Agronomy Department, PMAS Arid Agriculture University, Rawalpindi, Pakistan.
⁵ Global Food Systems Institute, University of Florida, Gainesville, FL, USA.

PMID: 38722337
DOI: 10.1007/s00484-024-02686-6

Abstract

Phenological shifts are one of the most visible signs of climatic variability and change in the biosphere. However, modeling plant phenological responses has always been a key challenge due to climatic variability and plant adaptation. Grapevine is a phenologically sensitive crop and, thus, its developmental stages are affected by the increase in temperature. The goal of this study was to develop a temperature-based grapevine phenology model (GPM) for predicting key developmental stages for different table grape cultivars for a non-traditional viticulture zone in south Asia. Experiments were conducted in two vineyards at two locations (Chakwal and Islamabad) in the Pothawar region of Pakistan during the 2019 and 2020 growing seasons for four cultivars including Perlette, King's Ruby, Sugraone and NARC Black. Detailed phenological observations were obtained starting in January until harvest of the grapes. The Mitscherlich monomolecular equation was used to develop the phenology model for table grapes. There was a strong non-linear correlation between the Eichhorn and Lorenz phenological (ELP) scale and growing degree days (GDD) for all cultivars with coefficient of determinations (R²) ranging from 0.90 to 0.94. The results for model development indicated that GPM was able to predict phenological stages with high skill scores, i.e., a root mean square (RMSE) of 2.14 to 2.78 and mean absolute error (MAE) of 1.86 to 2.26 days. The prediction variability of the model for the onset timings of phenological stages was up to 3 days. The results also reveal that the phenology model based on GDD approach provides an efficient planning tool for viticulture industry in different grape growing regions. The proposed methodology, being a simpler one, can be easily applied to other regions and cultivars as a predictor for grapevine phenology.

Keywords: Decision support tool; Developmental stages; Growth models; Heat units; Pothwar region; Viticulture.

Grants and funding

2AG4-029/Higher Education Commision, Pakistan