| Abstract |
Tree phenology is a major component of the global carbon and water cycle, serving as a fingerprint of climate change, and exhibiting significant variability both within and between species. In the emerging field of drone monitoring, it remains unclear whether this phenological variability can be effectively captured across numerous tree species. Additionally, the drivers behind interspecific variations in the phenology of deciduous trees are poorly understood, although they may be linked to plant functional traits. In this study, we derived the start of season (SOS), end of season (EOS), and length of season (LOS) for 3099 individuals from 74 deciduous tree species of the Northern Hemisphere at a unique study site in southeast Germany using drone imagery. We validated these phenological metrics with in-situ data and analyzed the interspecific variability in terms of plant functional traits. The drone-derived SOS and EOS showed high agreement with ground observations of leaf unfolding (R2 = 0.49) and leaf discoloration (R2 = 0.79), indicating that this methodology robustly captures phenology at the individual level with low temporal and human effort. Both intra- and interspecific phenological variability were high in spring and autumn, leading to differences in the LOS of up to two months under almost identical environmental conditions. Functional traits such as seed dry mass, chromosome number, and continent of origin played significant roles in explaining interspecific phenological differences in SOS, EOS, and LOS, respectively. In total, 55 %, 39 %, and 45 % of interspecific variation in SOS, EOS, and LOS could be explained by the Boosted Regression Tree (BRT) models based on functional traits. Our findings encourage new research avenues in tree phenology and advance our understanding of the growth strategies of key tree species in the Northern Hemisphere.
|
