*2.2. Field Sampling*

Birch (*B. ermanii*) forests are distributed between 1700 and 2200 m a.s.l. (upper limit) on the Changbai Mt. The *Betula* elevational gradient can be divided by the alpine treeline (2030 m a.s.l.) into two parts, i.e., below the alpine treeline with trees of > 2 m in height and above the alpine treeline with trees of < 2 m in height. We used this elevational gradient as a proxy of environmental change, mainly air warming [1,28]. To compare patterns of remobilization efficiency across environmental gradients from low elevations to high elevations, samples were taken at the peak-growing season and at the end of growing season. We selected end-July (23 July 2017) to collect samples, since for that time the difference in leaf phenology along the elevational gradient was negligible [29] and the physiological processes such as photosynthesis and growth are most active [30]. The end-season samples were collected from the higher elevations downslope to the lower elevations between September 5 and 9, when the nighttime temperature at an elevation dropped to below 0 ◦C according to a linear model using climate data from Tian-Chi Meteorological Station located at 2623 m a.s.l. and from the Meteorological Station (738 m a.s.l.) of Changbai Mountain Forest Ecosystem Research facility. Five healthy *B. ermanii* trees (*n* = 5) with similar age, height and diameter at breast height (DBH) or basal diameter (2107 and 2187 m a.s.l.) were selected at each of the six elevations, i.e., 1700, 1800, 1900, 2027, 2107 and 2187 m a.s.l. (Table 1). Using a temperature lapse rate of 0.6 ◦C decrease with 100 m increase in elevation, the annual mean air temperature is 3 ◦C lower at 2187 m than at 1700 m a.s.l. One-year-old shoots (with bark), main stem sapwood (with bark) and fine roots (< 0.5 cm in diameter, with bark) were collected from each selected tree. For leaf samples, mature green leaves were collected on 23 July (mid-season samples), whereas only freshly fallen leaves were collected from the forest floor close to each sample tree (end-season samples), to minimize difference in leaf phenology among elevations. Also, to minimize the effects of diurnal temperature variation and sunlight on samples, all samples were taken around noon and immediately stored in a cool box [31]. The samples were killed in a microwave oven at 600 W for 40 s to minimize the physiological activities, dried to a constant weight at 65 ◦C and then ground to fine powder to pass through a 1-mm sieve for analysis. We measured the height and DBH of each tree at the first sampling date (July) (Table 1). For juvenile trees with DBH of < 5 cm, basal diameter of trees (about 1 cm above ground surface) was measured in July. As limitations have been imposed on tree-harvesting within the Changbai Mountain Nature Reserve, we, therefore, only measured the biomass of leaves, shoots, stems, and roots of 5 average trees at the two highest elevations (2107 and 2187 m a.s.l.) in July 2017.


**Table 1.** Characteristics of the plots and the sampling trees *Betula ermanii* Cham. (mean ± 1 standard deviation; *n* = 5 trees) on the Changbai Mountain (Jilin, NE China).

a basal diameter (about 1 cm above ground surface).
