Wood is of crucial importance for man and biosphere. In this mini review, we present the fundamental processes involved in tree-ring formation and intra-annual dynamics of cambial activity, along with the influences of the environmental factors. During wood formation, new xylem cells produced by the cambium are undergoing profound transformations, passing through successive differentiation stages, which enable them to perform their functions in trees. Xylem cell formation can be divided in five major phases: (1) the division of a cambial mother cell that creates a new cell; (2) the enlargement of this newly formed cell; (3) the deposition of its secondary wall; (4) the lignification of its cell wall; and finally, (5) its programmed cell death. In most regions of the world cambial activity follows a seasonal cycle. At the beginning of the growing season, when temperature increases, the cambium resumes activity, producing new xylem cells. These cells are disposed along radial files, and start their differentiation program according to their birth date, creating typical developmental strips in the forming xylem. The width of these strips smoothly changes along the growing season. Finally, when climatic conditions deteriorate (temperature or water availability in particular), cambial activity stops, soon followed by cell enlargement, and later on by secondary wall deposition. Without a clear understanding of the xylem formation process, it is not possible to comprehend how annual growth rings and typical wood structures are formed, recording seasonal variations of the environment as well as extreme climatic events.

Key message: We demonstrate that, beyond leaf phenology, the phenological cycles of wood and fine roots present clear responses to environmental drivers in temperate and boreal trees. These drivers should be included in terrestrial ecosystem models.Context: In temperate and boreal trees, a dormancy period prevents organ development during adverse climatic conditions. Whereas the phenology of leaves and flowers has received considerable attention, to date, little is known regarding the phenology of other tree organs such as wood, fine roots, fruits, and reserve compounds.Aims: Here, we review both the role of environmental drivers in determining the phenology of tree organs and the models used to predict the phenology of tree organs in temperate and boreal forest trees.Results: Temperature is a key driver of the resumption of tree activity in spring, although its specific effects vary among organs. There is no such clear dominant environmental cue involved in the cessation of tree activity in autumn and in the onset of dormancy, but temperature, photoperiod, and water stress appear as prominent factors. The phenology of a given organ is, to a certain extent, influenced by processes in distant organs.Conclusion: Inferring past trends and predicting future trends of tree phenology in a changing climate requires specific phenological models developed for each organ to consider the phenological cycle as an ensemble in which the environmental cues that trigger each phase are also indirectly involved in the subsequent phases. Incorporating such models into terrestrial ecosystem models (TEMs) would likely improve the accuracy of their predictions. The extent to which the coordination of the phenologies of tree organs will be affected in a changing climate deserves further research.

Our knowledge about the influences of environmental factors on tree growth is principally based on the study of dominant trees. However, tree social status may influence intra-annual dynamics of growth, leading to differential responses to environmental conditions. The aim was to determine whether within-stand differences in stem diameters of trees belonging to different crown classes resulted from variations in the length of the growing period or in the rate of cell production. Cambial activity was monitored weekly in 2006 for three crown classes in a 40-year-old silver-fir (Abies alba) plantation near Nancy (France). Timings, duration and rate of tracheid production were assessed from anatomical observations of the developing xylem. Cambial activity started earlier, stopped later and lasted longer in dominant trees than in intermediate and suppressed ones. The onset of cambial activity was estimated to have taken 3 weeks to spread to 90 % of the trees in the stand, while the cessation needed 6 weeks. Cambial activity was more intense in dominant trees than in intermediate and suppressed ones. It was estimated that about 75 % of tree-ring width variability was attributable to the rate of cell production and only 25 % to its duration. Moreover, growth duration was correlated to tree height, while growth rate was better correlated to crown area. These results show that, in a closed conifer forest, stem diameter variations resulted principally from differences in the rate of xylem cell production rather than in its duration. Tree size interacts with environmental factors to control the timings, duration and rate of cambial activity through functional processes involving source–sink relationships principally, but also hormonal controls.