B051-0011
Importance of Photoperiod in Process-based Models for Predicting the Spring Flowering Phenology of Temperate Species in the Era of Climate Change

Thursday, 10 December 2020
Poster
Sukyung Kim1, Tae Kyung Kim1, Sukhee Yoon2, Keunchang Jang2, Myoungsoo Won3, Jonghwan Lim2 and Hyun Seok Kim4,5, (1)Seoul National University, Seoul, Korea, Republic of (South), (2)National Institute of Forest Science, Forest Ecology and Climate Change Division, Seoul, South Korea, (3)National Institute of Forest Science, Forest Resource Management and Restoration Division, Seoul, South Korea, (4)Seoul National University, Department of Forest Sciences, Seoul, South Korea, (5)Seoul National University, Research Institute for Agriculture and Life Sciences, Seoul, South Korea
Abstract:
Usage of process-based models for the prediction of flowering time over simple statistical models has been increased due to the climate change and recent frequent extreme weather events. Therefore, validations of various process-based models are required for variety of species because of the complexity of flowering mechanisms. In this study, we developed and evaluated six process-based models to predict the flowering time of seven spring flowering species (Cornus officinalis, Forsythia koreana, Rhododendron mucronulatum, Prunus yedoensis, Rhododendron yedoense var. poukhanense, Rhododendron schlippenbachii, Robinia pseudoacacia) using phenological data observed from 10 different arboretums in South Korea over 9 years. Generally, the Growing Degree Days (GDD) model showed better prediction performance than the sequential chilling-forcing model, largely owing to uncertainties in the estimation of chilling requirements in the chilling-forcing model and large spatial variations in winter temperatures. Additionally, the prediction accuracies of all models improved with the inclusion of the photoperiod as an additional parameter. This improvement in performance was different depending on the model and the chilling-forcing model showed greater improvement than the GDD model. Our analysis revealed that the photoperiod reduced the effects of chilling requirements and starting date of forcing accumulation of models, while it emphasized the importance of forcing requirement. As a result of comparing the effect of photoperiod on flowering time by species, Cornus officinalis and Robinia pseudoacacia, blooming at the earliest and the latest, respectively, among the seven species were less responsive to photoperiod than the other five species. Our study showed the differences in the importance of parameters required for process-based models depending on the flowering time of species in temperate regions.
<< Previous Abstract | Next Abstract >>