1. Temperate trees rely on carbon (C) and nutrient remobilisation from storage to resume growth after winter. Minimum storage levels during the growing season suggest that remobilisation could signify that C availability is insufficient to meet growth demands; consequently, growth might be C and/or nutrient limited. However, it remains unclear whether higher growth demands are covered by higher remobilization. This study examined whether higher C and nutrient demands associated with fast growth or deciduousness, rely on greater remobilisation.

2. In 11 sympatric deciduous and evergreen angiosperm tree species from southern South América, the magnitude of seasonal remobilisation of C and nutrient storage was assessed as the seasonal minimums (relative to seasonal maximums) of whole tree non-structural carbohydrates (NSC), nitrogen (N), and phosphorus (P) concentrations and pools. The basal area increment and stem wood density were determined for each tree, from which the biomass radial increment (BRI) was estimated. The effects of leaf habit and BRI on the seasonal minimums NSCs and nutrient concentrations and pools were analysed using linear mixed-effects models.

3. Radial growth was not related to seasonal minimum NSC or nutrient concentrations and pools in either the evergreens or deciduous angiosperms; thus, faster growth was not associated with greater remobilisation of C or nutrient stores. Further, larger trees grew faster than smaller ones, but did not have higher remobilisation. Deciduous species had higher year-round whole tree NSC and nutrient concentrations than evergreens; however, both groups had similar BRI and seasonal minimum concentrations and pools of NSCs and nutrients.

4. Neither growth rate nor leaf habit drove the magnitude of C and nutrient remobilisation in the angiosperm trees examined here, indicating no C or nutrient limitation. This result contradicts the view that growth and storage strongly regulate one another, as proposed by a growth-storage trade-off.

Seasonal sampling of the dominant tree species of four cold-temperate forests was performed in Patagonia, specifically in the Aysén region of Chile. The sites were: Aiken (45°27´S, 72°44´W, 45 m a.s.l.), Coyhaique (45°32´S, 72°03´W, 466 m a.s.l.), Tranquilo (46°38´S, 72°47´W, 331 m a.s.l.) and Exploradores (46°29´S, 73°09´W, 137 m a.s.l.).

In each forest, six unshaded and healthy juvenile trees of two winter deciduous species (Nothofagus antarctica and Ribes magellanicum, Na and Rm, respectively) and two to five evergreen angiosperm species (Nothofagus nitida (Nn), Embothrium coccineum (Ec), Aristotelia chilensis (Ac), Myrceugenia planipes (Mp), Luma apiculata (La), Nothofagus dombeyi (Nd), Rhaphithamnus spinosus (Ra), Nothofagus betuloides (Nb), Drimys winteri (Dw)) were selected for chemical analyses and to measure radial growth. Coarse roots, stems, branches, twigs, and new leaves were sampled between 10:00 and 16:00 h at a single time point (hereafter referred to as “sampling date”), which corresponded with mid-spring, mid-summer, mid-autumn, and mid-winter. For each sampling date, the four forests were sampled over a 3-week period.

One to-the-pith long stem core at breast height (1.35 m) was sampled using a 5.15 mm increment borer for NSC analyses. Another stem core was similarly sampled on the first sampling date to determine tree growth. Growth was measured as the increment in basal area of the last 6 years (BAI₆). To account for potential differences in biomass investment due to wood density, BAI₆ was converted into the biomass radial increment (BRI), which represents the biomass increment of a hollow cylinder of 1 cm height, for which the wall thickness (i.e. the difference between outer and inner diameter) is the BAI₆. The volume of this cylinder was then multiplied by the species wood density of a given species to obtain the biomass of the cylinder.

All samples for chemical analyses were placed in a forced-air stove at 65 °C for 72 h. The material was then ground into a fine powder and stored under cool, dry, and dark conditions, until chemical analyses were conducted.

Carbohydrate concentrations were analysed for all the samples. C reserves were assessed by determining NSC; specifically, as the sum of the three most abundant low molecular-weight soluble sugars (glucose, fructose and sucrose; hereafter SGF) and starch. NSC concentrations were analysed following the procedure of Hoch et al. (2002) with some modifications as described by Piper et al. (2019). Nitrogen (N) concentrations were determined from 0.2 g of dried ground material using a combustion analyser (LECO TruSpec Micro CHN, Centro de Investigación de Ecosistemas de la Patagonia, Coyhaique, Chile). Inorganic total phosphate (P) concentrations were determined from 100 mg dried ground material based on the procedure described by Murphy & Riley (1962).

The dry wood biomass of each organ was estimated using previously developed allometric functions. For each sampled tree and on each date, the NSC, starch, SGF, N, and P pools were first calculated per each organ as the product between the biomass and the concentration for each organ. Then, the pools of the different organs of a tree were summed to estimate the whole-tree total pool on each sampling date. The seasonal minimum pool of NSC, starch, SGF, N, and P was determined for each tree on the four sampling dates (as the lowest value), and was expressed relative to the seasonal maximum pool (percentage). Thus, the higher the seasonal minimum, the lower the amplitude of seasonal variation in a given pool (Fig. 1). Similar calculations were performed for NSC, starch, SGF, N, and P concentrations; however, in these cases, seasonal minimums were estimated separately for each organ.

There are some missing values of N and P concentrations.

Wood density was measured for most sites in this study, although for a few cases (species in a given site) I used published data available elsewhere.