The majority of terrestrial primary production is performed by plants, the ontogenetic growth trends of which greatly influence biomass and carbon dynamics. Here, we study ontogenetic trends of primary (apical) and secondary (stem thickening) growth in Arctic (Svalbard, Norway) and alpine (Krkonoše, Czechia) populations of black crowberry (Empetrum nigrum), the dominant plant species of certain tundra communities. The environmental conditions in alpine areas are more favourable for plant growth than those in the High Arctic, where temperatures are lower, there is less precipitation and soils are shallower, among other differences. These differences were reflected in significant differences in absolute growth rates and shrub age between the populations under study. However, we found almost no differences in ontogenetic growth trends between the populations. In both populations, primary growth and secondary (stem base) growth decrease throughout ontogeny whereas secondary (stem top) growth and basal area increment increase. No significant differences in the slope of the trends were found in either primary or secondary (stem base) growth. Trends in the ratio between basal area increment and primary growth revealed neither absolute nor relative differences between the populations. Ontogenetic trends in the shrubs analysed were surprisingly stable despite the prominently different environmental conditions. Empetrum plants have adapted to the different environments by altering their absolute growth rate only. This adaptation has probably also resulted in the different longevity of plants constituting the study populations, confirming the theory that slower-growing plants live longer. Primary growth and secondary (diameter) growth at the stem base seem to be more basic characteristics of plant growth compared to basal area increment and secondary (diameter) growth at the apex because the latter two seem to be dependent on the absolute growth rate.

Study sites and sampling

Arctic and alpine populations of the black crowberry (Empetrum nigrum L.), be it the nominal subspecies or E. nigrum subsp. hermaphroditum (Hagerup) Böcher, is exposed to different climatic, soil, and vegetation conditions. We sampled forty individuals from an Arctic population at the end of August 2018 in the Colesdalen valley (78°6′ N; 15°5′ E), Svalbard, Norway. The average annual precipitation in the valley is 190 mm, mostly in the form of snow. From 1969 to 1990, the average annual temperature was −6.7°C, the coldest month was February, with an average temperature of −16.2°C, and the warmest month was July, with an average temperature of 5.9°C (Longyearbyen Airport station, 18 km north of the site; Norwegian Meteorological Institute, 2010). The onset of the growing season spans between days number 165 and 174 (12–23 June) in the Svalbard Islands. Samples were taken at elevations between 40 to 80 m a.s.l. on a south-facing slope with an inclination of 10–20%. Shallow and poor soils were present on the site. There were no other shrubs or herbs growing in the vicinity of the individuals sampled.

The temperate alpine population was represented by 41 individuals collected in July and August 2019 in the Krkonoše Mountains, Czechia. Both co-occurring subspecies of E. nigrum (subsp. nigrum and subsp. hermaphroditum) were sampled because they are impossible to identify in the field. Individuals were sampled at three localities: Zadní Planina (50°42′49″ N; 15°40′35″ E), Pančavská louka (50°45′56″  N; 15°32′38″ E), and Labská louka (50°46′12″ N; 15° 32′20″ E). All samples were taken at elevations between 1,300 and 1,400 m a.s.l. The locality at Zadní Planina was on a 20% steep south-eastern slope and the other two localities were flat. At the elevation of 1,603 m a.s.l., the average annual temperature is 0.2°C, the warmest month being July (8.3°C) and the coldest January (−7.0°C). The average annual precipitation reaches 1,400 mm. The onset of cellular division (i.e. the growing season) in co-occurring Pinus mugo is between days 138 and 145 (18–25 May) in the Krkonoše Mts. Region is botanically relatively well described and Krkonoše Mts show high plant endemism from the central European perspective. Sampled shrubs occurred in closed vegetation with Calluna vulgaris, Vaccinium myrtillus, Vaccinium uliginosum, Pinus mugo, and several grass species (Poaceae).

The sampling consisted of the extraction of complete individuals (in clonal alpine individuals only the relevant part), including lignified roots (to allow the laboratory assessment of radial growth and plant age from the root collar) and whole aboveground organs. Root samples were cleaned, placed in a marked plastic bag, and treated with 40% alcohol to prevent mold.

Growth measurements

The method of serial sectioning was used in order to determine the rate of primary growth. It consisted of tree-ring measurements of cross-sections at different heights and cross-dating the obtained series at the intra-individual level. The first cross-section was always made at the root collar or the bottom part of the stem and was followed by one to ten cross-sections evenly spaced across the shrub body so that the distances between the cross-sections were approximately 10 cm. All branches longer than 10 cm were included to account for the species’ greater growth complexity, as it branches heavily. The exact distances between all the cross-sections were measured for subsequent primary growth analysis.

Slices no thicker than 20 µm were extracted from all cross-sections using a sledge microtome. The slices were stained using a 1:1 Safranin/Astra blue solution and embedded in Canada balsam. The entire area of each slice (i.e. whole cross-section) was photographed at 100× magnification in order to measure annual rings along multiple axes as well as to detect partially missing annual rings and other growth defects. Measurements were obtained from the images using NIS-Elements software for two radii per cross-section. The first measurement was always performed for the longest radius whereas the second measurement, aimed to reveal partially missing rings, was made at an angle of at least 90° to the first radius. Cross-dating was done mostly visually using PAST5 software, first at the cross-sectional level by comparing radii measurements and, secondly, at the level of the whole individual (including branches) by comparing mean cross-sectional series. In total, we were able to use a series from 30 individuals from Svalbard (92 missing and 232 partially missing annual rings detected) and 33 individuals from the Krkonoše Mts (23 missing and 126 partially missing annual rings detected). The remaining samples were not cross-dated successfully and were excluded from further analyses.

We calculated the primary growth rate between cross-sections using the difference in the number of tree rings and stem length between two consecutive cross-sections for each considered branch. For the corresponding set of tree rings, we also calculated the rate of secondary growth at the shrub base (using measurements from the basal cross-section) and at the shrub top (using the measurements at the lower cross-section of each pair). We also calculated secondary growth expressed as basal area increment (measured in mm²) and the ratio between basal area increment and primary growth (measured in mm). Basal area increment, in contrast to simple ring width, includes the circumference of the stem on which the ring is produced and is, therefore, assumed to be more representative of stem conducting capacity and stem biomass produced.

Statistics

To test for relationships between the growth variables (primary growth, secondary growth at the base and top, basal area increment, and growth ratio), shrub age, and population age structure, we used multiple linear and polynomial regression models. These models were based on average growth rates within shrub sections. We also used a subset of the data for shrubs younger than ten years to avoid bias caused by large differences in shrub age between the two study regions. Similarly, to remove the age effect from the relationship, the residuals of growth vs age models were used to test for a relationship between primary and secondary growth. Finally, we calculated the age distribution of each sample population and tested for differences in these distributions between the populations. All the analyses were performed in R (R Core Team 2020).