Abstract: Urbanization-induced warming advanced the timing of spring budburst, impacting on urban ecosystems. However, how urban artificial light affects the spring budburst and its spatial variation within species distribution are less studied, especially lacking experimental evidences. Here, we conducted a climate-controlled experiment using twigs collected from artificial light (AL) and no-artificial light (NoAL) conditions at three latitudinal gradients (L_high, L_middle and L_low) in China. We found that the temperature responsiveness of spring budburst (T_res, defined as the number of days to budburst after the twigs are placed into the chambers, with a smaller value indicating stronger responsiveness) was significantly stronger for NoAL individuals (54.3 days) than AL individuals (60.7 days). Additionally, AL twigs exhibited a greater photoperiod limitation (12.7 days vs. 7.6 days) and a higher heat requirement (732.15 K vs. 679.15 K) than NoAL twigs, suggesting that individuals exposed to artificial light may have adapted to longer photoperiod and increased the heat requirement for budburst. More importantly, T_res difference between AL and NoAL individuals was more pronounced in northern sites (5.8 days at L_high, 12.2 days at L_middle) than in southern sites (0.7 days at L_low), possibly due to higher inter-annual temperature variability at higher latitudes. Our findings provide experimental evidence of the effect of artificial light on tree budburst and highlight the need to consider the adaptability of urban trees when studying phenological responses to climate change in urban environments.

Key words: urbanization, artificial light, spring phenology, photoperiod, spatial variation