Photoperiod drives cessation of wood formation in northern conifers

Aims: Shifts in xylem phenology directly determine the forest capacity for carbon sequestration. However, a systematic understanding of the spatial patterns and the underpinning drivers in determining the cessation of wood formation (C_cw) is lacking at a pan-continental scale. Here, we addressed this knowledge gap by compiling a new dataset of multiple xylem phenology timings for northern conifers. Locations: Sixty-two study sites, Northern Hemisphere (25–55° N). Time period: 2003–2018 (16 years). Taxa: Thirty-three conifer species. Methods: A generalized additive model was fitted to characterize the latitudinal pattern in C_cw. Structural equation modelling and a linear mixed-effects model were applied to determine the main drivers underlying the latitudinal pattern in C_cw. Results: The C_cw followed a flat S-shaped pattern with increasing latitude. Photoperiod was the dominant determinant of the latitudinal pattern of C_cw, and a longer photoperiod was associated with an earlier C_cw. Both mean growing-season temperature and total growing-season precipitation exhibited significantly positive relationships to the cessation of cell elongation and thus the C_cw across all study sites. In arid regions, the pre-growing-season temperature had a significantly negative effect on C_cw. In humid regions, C_cw was positively affected by the mean growing-season temperature. The onset of wood formation showed significantly positive coupling with C_cw at arid sites but not at humid sites. Early successional species were sensitive to hydrothermal variations during the pre-growing season. Main conclusions: We reveal the dominant role of photoperiod in determining the cessation of wood formation for northern conifers and highlight differentiated interactive effects between photoperiod and seasonal climatic factors and the preceding xylem phenophases in determining C_cw among ecoregions and tree species. These insights provide evidence to reduce uncertainty in prediction of the forest carbon uptake potential and the consequent biophysical feedbacks of northern forests.