Root diameter, a critical indicator of root physiological function, varies greatly among tree species, but the underlying mechanism of this high variability is unclear. Here, we sampled 50 tree species across tropical and temperate zones in China, and measured root morphological and anatomical traits along the first five branch orders in each species. Our objectives were (i) to reveal the relationships between root diameter, cortical thickness and stele diameter among tree species in tropical and temperate forests, and (ii) to investigate the relationship of both root morphological and anatomical traits with divergence time during species radiation. The results showed that root diameter was strongly affected by cortical thickness but less by stele diameter in both tropical and temperate species. Changes in cortical thickness explained over 90% of variation in root diameter for the first order, and ∼74–87% for the second and third orders. Thicker roots displayed greater cortical thickness and more cortical cell layers than thinner roots. Phylogenetic analysis demonstrated that root diameter, cortical thickness and number of cortical cell layers significantly correlated with divergence time at the family level, showing similar variation trends in geological time. The results also suggested that trees tend to decrease their root cortical thickness rather than stele diameter during species radiation. The close linkage of variations in root morphology and anatomy to phylogeny as demonstrated by the data from the 50 tree species should provide some insights into the mechanism of root diameter variability among tree species.

Tree roots are highly heterogeneous in form and function. Previous studies revealed that fine root respiration was related to root morphology, tissue nitrogen (N) concentration and temperature, and varied with both soil depth and season. The underlying mechanisms governing the relationship between root respiration and root morphology, chemistry and anatomy along the root branch order have not been addressed. Here, we examined these relationships of the first- to fifth-order roots for near surface roots (0–10 cm) of 22-year-old larch (Larix gmelinii L.) and ash (Fraxinus mandshurica L.) plantations. Root respiration rate at 18 °C was measured by gas phase O₂ electrodes across the first five branching order roots (the distal roots numbered as first order) at three times of the year. Root parameters of root diameter, specific root length (SRL), tissue N concentration, total non-structural carbohydrates (starch and soluble sugar) concentration (TNC), cortical thickness and stele diameter were also measured concurrently. With increasing root order, root diameter, TNC and the ratio of root TNC to tissue N concentration increased, while the SRL, tissue N concentration and cortical proportion decreased. Root respiration rate also monotonically decreased with increasing root order in both species. Cortical tissue (including exodermis, cortical parenchyma and endodermis) was present in the first three order roots, and cross sections of the cortex for the first-order root accounted for 68% (larch) and 86% (ash) of the total cross section of the root. Root respiration was closely related to root traits such as diameter, SRL, tissue N concentration, root TNC : tissue N ratio and stele-to-root diameter proportion among the first five orders, which explained up to 81–94% of variation in the rate of root respiration for larch and up to 83–93% for ash. These results suggest that the systematic variations of root respiration rate within tree fine root system are possibly due to the changes of tissue N concentration and anatomical structure along root branch orders in both tree species, which provide deeper understanding in the mechanism of how root traits affect root respiration in woody plants.

Despite extensive research on the seasonal dynamics of non-structural carbohydrate (NSC) and nitrogen (N) concentrations, the size and relative contributions of NSC and N pools across different tree organs are not well understood. We have measured the changes in NSC and N concentrations in leaves, branches, stems and all root branch orders at monthly intervals in control and girdled trees of larch (Larix gmelinii) and ash (Fraxinus mandshurica). The biomass of each plant compartment was also determined to calculate the size of the NSC and N pools. In both species, 13–37 % of the NSC and N pools were mobilized at the beginning of the growing season. Among the mobilized pools, stems and non-absorptive roots (branch orders 4–9) acted as the largest NSC sources in larch and ash, respectively, while branches served as the largest N source in both species. After stem girdling, 22 and 50 % of the root NSC stores in larch and ash, respectively, were mobilized to maintain root activities during the growing season. Tree mortality was observed 1 year after girdling, at which time there was still an abundant NSC pool in the roots. We conclude that (1) different storage organs differ in their contribution to new tissue growth at the beginning of the growing season and that those storage organs holding higher fractions of the NSC or N pool are not necessarily those which mobilize more NSC or N; (2) tree growth may not be limited by carbon (C) availability; (3) C storage in non-absorptive roots plays an important role in maintaining tree survival after the termination of photosynthate flow from aboveground sources.