Climate change including variations in temperature, precipitation and solar radiation have significantly affected the phenology, the growth duration, and thus the crop yield, although large uncertainties remain in terms of impact magnitude, spatial pattern and mechanisms (Ortiz et al., 2008; Tao et al., 2013; Y. Chen et al., 2018). The ongoing increase in carbon dioxide would result in further global warming. Since 1960, the annual average temperature has increased by 1.2°C in China, and is estimated to continuously increase by 1 to 5°C by 2100 (Solomon, S. et al., 2007; Piao et al., 2010). At the same time, the drier regions in North China are receiving less precipitation, while the wetter regions in South China are experiencing more rainfall, which directly induce a more significant spatial difference of precipitation (Guo et al., 2020; Xin et al., 2020). Moreover, the climate extremes, especially drought and flood, appeared frequently in the past few decades, exerting further ecological and agricultural pressure (Deryng et al., 2014; Hatfield and Prueger, 2015; R. Chen et al., 2018). With the increase of greenhouse emissions, the trend of climate warming will continue in the coming decades, and will be more prominent in cultivated areas (Lobell et al., 2011), which would bring higher risk for future regional food security (Ainsworth et al., 2008; Leakey, 2009; Vanuytrecht et al., 2012; Pugh et al., 2016). The heat accumulation during growth period, which is usually characterized by Growing Degree Days (GDD), is closely controlled by climate change, directly related to the energy needs for crop development, and further exert a significant influence on crop yield. Besides, the nitrogen deficiency, water shortage, and photoperiod requirements should not be ignored for vegetation development. Researches using either environment-controlled experiments (Ottman et al., 2012), or historical records (Lobell et al., 2011; Tao et al., 2014), or crop model simulations (Porter et al., 2015; Asseng et al., 2013) have documented that global warming would shorten the crop growth duration and then reduce crop yields (Porter et al., 2015). However, for area with insufficient heat such as temperate areas, raising temperature could complement heat defect and thus lead to the increase of yield (Tao et al., 2013; Tao et al., 2014; Zhang et al., 2014). Therefore, given the significance of GDD to agricultural productivity and its regional uncertainty, understanding the GDD variations and their driving factors is crucial to agriculture monitoring and management (Yin et al., 2019).
Phenology determines the crop growth period. For the whole life cycle of early rice, growth period begins in spring with the photosynthesis starting when temperature and solar radiation increase up to a certain threshold (spring phenology, transplanting for rice), and terminate in autumn as photosynthesis ceases when temperature and solar radiation decrease (autumn phenology, maturity for rice) (Kimball et al., 2004; Euskirchen et al., 2006). As the metric of heat requirement for growth and development, GDD is controlled by both the growth duration and temperature, and is mainly affected by both phenology and climate change (Deng et al., 2018). Previous studies indicated that GDD is positively related to the seedling recruitment in the context of warming near the boreal forest (Miller et al., 2017). The GDD of agriculture system, however, is relatively complex due to the mixed effects by both climatic and agronomic factors, which remains great uncertainty under changing climate (Hildén et al., 2005; Estrella et al., 2007). For example, an increase in daily atmospheric temperature leads to more rapid increase in degree days and hasten maturity, finally resulting in lowered biomass and decreased wheat yield (Zhao et al., 2007; Hatfield and Prueger, 2015; Aslam et al., 2017; Prasad et al., 2017). In the U.S. Corn Belt, after removed the effects of cultivar shift, the number of GDD needed for corn progress was declined, and the yields would have been 12.6 bu ac− 1 lower around 2005 (Sacks and Kucharik, 2011). However, studies focusing on GDD of early rice, one of the most widely planted crops around the world, are scarce. Besides, GDD in agricultural researches were usually used to reflect regional heat resources, while individual-specific actual heat accumulation during growth period is always ignored, and therefore need to be discussed. Since cultivated area accounts for about 12% of the surface area (Ramankutty and Foley, 1998), shifts in crop growth period and GDD would exert great influence on water and carbon fluxes, and thus regional and even global climate (White et al., 1999; Baptist and Choler, 2008; Wu et al., 2012; Tao et al., 2012). Furthermore, GDD directly impose influence on crop yield and productivity, which consequently bear great significance for food security and economic stability (Challinor et al., 2010; Asseng et al., 2011; Tao et al., 2012; Licker et al., 2013). Therefore, in this study, we choose early rice, one of the most important part of agriculture in China, which is directly related to national economy and people’s livelihood, to investigate the variability of heat accumulation under climate change, and its driving factors.
Crop GDD is mainly regulated by both temperature and growth duration. On the one hand, climate change and agricultural measures directly affect the heat accumulations of crop by changing the temperatures during growth period; on the other hand, they induce shifts in phenology, which influences the timing and duration of a photosynthetically active canopy and changes of heat accumulation during growth period ensued (Keeling et al., 1996; Myneni et al., 1997). In this analysis, based on in situ agrometeorological experimental data of early rice across 1981 to 2010 in China, we discuss the main factors affecting GDD and its influencing mechanism for early rice. The primary objectives of this study were: (1) to explore the changes of phenological dates and growing season over the past three decades, (2) to investigate the tempo-spatial patterns of GDD changes during each of the growing periods, (3) to identify the relationships between GDD and phenological dates using partial correlation analyses, (4) to determine the main factors and mechanisms leading to changes in heat accumulations for each of the growing periods.