Research on how plant ecological strategies (competitive, stress-tolerant, or ruderal) vary within species may improve our understanding of plant and community responses to climate warming and also successional changes. With increasing temperature, the importance of ruderal (R) and stress-tolerance (S) components is hypothesized to decrease, while the strength of the competitive (C) component should increase. Offshoots and younger plants are predicted to have greater R and smaller S components. Leaf area, leaf dry matter content, and specific leaf area were measured for 1,344 forest plants belonging to 134 species in Northeastern China, and C, R, and S scores calculated for each. Linear mixed effect models were used to assess how these indicators differed among study sites (n = 2), regeneration types, ontogenetic stages, and plant life forms. The two study sites have an average annual temperature difference of 0.675°C, simulating a temperature increase of 0.630°C due to climate warming. Higher temperatures reduce low-temperature stress and frost damage, which may explain the observed decrease in R and S scores; at the same time, plant competitive ability increased, as manifested by higher C scores. This effect was most pronounced for herbs, but nearly negligible as compared to the effect of regeneration type for trees and of ontogeny for woody species. Resprouting trees and younger woody plants had higher R scores and lower S scores, a sign of adaptation to high disturbance. A small increase in mean annual temperature led to shifts in CSR strategy components for herbaceous species, without altering the vegetation type or community composition. Offshoots and younger plants had higher R and lower S scores, shedding light on similar changes in the ecological strategies of tree communities during secondary succession, such as the transition of Quercus mongolica coppices to forest, and age-related changes in Populus davidiana-Betula platyphylla forests.

The leaves of all tree, shrub, and herbaceous species (A total of 134 species, including 25 trees, 28 shrubs, and 81 herbs, 123 species in FL and 120 in LS) were extensively sampled according to tree regeneration types and woody plant ontogenetic stages. For each type, the goal was to obtain at least six replicas per site; however, for a few rare plant species, only three to five replicates were possible. Very few plant species were too rare to be sampled sufficiently. In order to ensure enough regeneration samples (i.e., at least six replicates of each regeneration type per site) and to accurately identify resprouting versus seed regeneration, only five tree species were sampled (A. sibirica, Acer mono, B. platyphylla, Padus racemose, and Q. mongolica); additionally, sampled individuals including seedling and resprouting were mature trees that had flowered and fruited. For trees, four ontogenetic stages were established: seedling, sapling, young adult, and adult. Seedlings were defined for coniferous trees as individuals with height < 30 cm and for broad-leaved species as individuals with height < 50 cm. Coniferous saplings had heights > 30 cm and broad-leaved saplings heights > 50 cm, in addition to a diameter at breast height (DBH) < 5 cm. Young adult trees were those that had just entered the reproductive stage, and generally had DBH ranging from 5 to 15 cm. Finally, adult trees were reproductively mature and had DBH > 30 cm. For shrubs, two ontogenetic stages were recognized: saplings and adults. Shrub saplings were structurally simple, unbranched individuals of height > 50 cm that showed an absence of reproductive organs, while adults had entered the reproductive stage.

The leaf area (LA), leaf dry matter content (LDMC), and specific leaf area (SLA) were determined according to Pérez-Harguindeguy et al. (2013). Leaf sampling took place from July 19 to 23, 2019 in LS, and from July 25 to 28, 2019 in FL; at this time, the leaves had fully expanded and reached maturity. Three to five unshaded directions were selected in the canopy for sampling, and vigorously growing branches were then collected for leaf extraction. Tall trees were sampled by professional climbers (Figure S1). Three to thirty undamaged, fully expanded leaves were measured per sampled individual. The C, S, and R scores for each individual were calculated using “StrateFy” (Pierce et al., 2017).