Spatial differences in the response of Ginkgo spring leaf phenology to climate change

Climate change-induced spring phenological shifts exert significant effects on terrestrial carbon, water and nutrient cycles as well as the climate system. Spring leaf-out of temperate trees is triggered by a combination of winter chilling, spring temperature and photoperiod. Yet, the effects of these cues along elevation and latitude remain poorly understood due to the lack of experiments. Here we test for elevational and latitudinal differences in the responses of spring leaf phenology to climate by conducting a manipulative experiment with four temperature (5, 10, 15, and 20 °C) and two photoperiod (8 and 16 h) treatments, using twigs from mature Ginkgo trees at two latitudes in China (39° 54′ N and 30°19′ N) and three elevations at the low latitude site (344, 826, 1098 m). We found that the responsiveness of leaf unfolding to temperature and photoperiod decreases with increasing elevation. Specifically, the temperature sensitivity of leaf unfolding (defined as the leaf-out advance per degree warming, S_T) was higher (4.17 days °C⁻¹) and the photoperiod effect on S_T was larger (decrease of 1.15 days °C⁻¹) at the lowest elevation than at the highest elevation (S_T = 3.26 days °C⁻¹; decrease of 0.48 days °C⁻¹). This may be related to adaptation to local environments and self-protection mechanisms of trees at high elevations to avoid frost damage. In addition, we found that photoperiod only affected leaf unfolding at low latitudes, with long (16-h) days advancing leaf unfolding in low-latitude individuals by, on average, 8.1 days relative to short (8-h) days. Field phenological observations supported the experimental results. Our results suggest that, for a given species, populations growing under different climate conditions may exhibit different phenological responses to climate change, with individuals in warmer regions likely becoming increasingly limited by photoperiod as the climate warms further. We thus provide empirical evidence of spatial differences in the relative effects of spring temperature and photoperiod on spring leaf phenology, which is not yet accounted for in models of spring phenology.