Advances in research on fruit tree phenology

Abstract: Fruit tree phenology has aroused public and scientific attention due to the growing evidence that the timing of development stages is largely dependent on environmental factors. Particularly, as fruit tree phenology events are highly sensitive to climate fluctuations, the timing of these events has been used as the most salient and sensitive bioindicators of climate change. Studies have reported shifts in fruit tree phenology due to warming temperatures, leading to earlier spring and later autumn phenology, with potential impacts on the safety of fruit production. Modelling, assessing and monitor- ing of fruit tree phenological dynamics are therefore key requirements to improve the understanding of how fruit tree responds to global warming. In order to improve the understanding of fruit tree phenolo- gy, this paper has systematically reviewed important progresses in researches on fruit tree phenology, fo- cusing on fruit tree phenology and climate warming, the driving factors on phenological phenomena and phenology monitoring methods. Then, we pointed out some main research aspects in the near future. Among them, previous studies have indicated that changes in spring phenology of fruit crops as a result of increased temperature may manifest as advancement, delay or no change. These inconsistent re- sponses are due to the unbalanced relationship between forcing and chilling effects, i.e. variation in chilling effects due to winter warming and changes in forcing effects related to spring warming. The driving factors of main fruit tree phenological change are environmental conditions (air temperature, hu- midity, etc.) and agricultural management, and thus a comprehensive understanding of the driving fac- tors affecting fruit tree phenology can benefit to fruit production. Furthermore, to better understand the chilling effects of winter temperature and forcing effects of spring temperature on fruit tree spring phe- nology, several models have been developed to calculate chilling and forcing accumulation. For exam- ple, results from the Chill Overlap model were better than those from Sequential model on evaluating apple flowering time. Of the Chill Overlap models, those fitted with Triangular or Dynamic Chill model and the Growing Degree Hour (GHD) heat sub-model seem to have more biological rationale and per- formed well statistically. Traditional fruit tree phenological observations were based on field records, which can obtain the key biological events on a specific date. However, such observations can be time consuming and labor intensive to cover many species across a region and establish long term data re- cords. To timely and accurately measure fruit tree phenological events over different spatial scales, the remote sensing has emerged as a new and valuable tool. Recent studies have reported that the remote sensing can monitor fruit tree phenology by an array of ground, near-surface and orbital sensors, each with some advantages and disadvantages. Remote sensing can achieve qualitative analysis and quantify- ing analysis for crops’main phenological stages. Meanwhile, time-series vegetation indexes are used to qualitatively analyze crop phenology and the most commonly used methods are VI thresholding, curve- fitting, inflection point and maximum slope. Currently, use of remote sensing to quantify the key pheno- logical periods of fruit tree has mainly focused on flowering and maturity period. Some pioneering stud- ies using remote sensing methods to quantify fruit tree flowering phenology were based on the linkages of flower pigments and colors with spectral signatures and the differentiating spectral characteristics of flowers and other objects, which had a good example in almond floral phenology. Other researchers us- ing near-infrared reflectance predicted the optimal harvest date of fruits based on the information about the maturity, including fruit skin color and the varying internal fruit characteristics. In the future, (1) More relevant driving factors should be considered into fruit tree phenological models. (2) The investi- gation on molecular mechanism to fruit tree phenological changes should be strengthened. (3) Quantitative monitoring fruit tree phenology should be strengthened from new sensors at ground, near-surface and airborne level. (4) Ground observations, model simulations and remote sensing monitoring methods should be synergetically used for fruit tree phenology verification.