Extent of intraspecific trait variability in ecologically central and marginal populations of a dominant alpine plant across European mountains

Abstract

Background and aims: Studying trait variability and restricted gene flow between populations of species can reveal species dynamics. Peripheral populations commonly exhibit lower genetic diversity and trait variability due to isolation and ecological marginality, unlike central populations experiencing gene flow and optimal conditions. This study focused on Carex curvula, the dominant species in alpine acidic meadows of European mountain regions. The species is sparser in dry areas such as the Pyrenees and Balkans, compared to the Central-Eastern Alps and Carpathians. We hypothesized that distinct population groups could be identified based on their mean functional trait values and their correlation with the environment; we predicted that ecologically marginal populations would have stronger trait correlations, lower within-population trait variability (intraspecific trait variability, ITV) and lower genetic diversity than populations of optimal habitats.

Methods: Sampling was conducted in 34 populations that spanned the entire distribution range of C. curvula. We used hierarchical clustering to identify emergent functional groups of populations, defined by combinations of multiple traits associated with nutrient economy and drought tolerance (e.g. specific leaf area, anatomy). We contrasted the geographical distribution of these groups in relation to environment and genetic structure. We compared pairwise trait relationships, within-population trait variation (ITV) and neutral genetic diversity between groups.

Key results: Our study identified emergent functional groups of populations. Those in the southernmost ranges, specifically the Pyrenees and Balkan region, showed drought-tolerant trait syndromes and correlated with indicators of limited water availability. While we noted a decline in population genetic diversity, we did not observe any significant changes in ITV in ecologically marginal (peripheral) populations.

Conclusions: Our research exemplifies the relationship between ecological marginality and geographical peripherality, which in this case study is linked to genetic depauperation but not to reduced ITV. Understanding these relationships is crucial for understanding the biogeographical factors shaping trait variation.

Keywords: Alpine plants; European mountains; cluster regression; ecological marginality; functional biogeography; intraspecific trait variability; neutral genetic diversity.

Fig. 1.

(A) Distribution of the sampling locations across the temperate European mountains. Symbols have been colour coded based on geographical region. The dashed line separates the two main genetic lineages (Western and Eastern clades) according to Pușcaș et al. (2008a). Population details can be found in Supplementary Data Table S1. (B) Hierarchical clustering of the sites based on geographical distances. (C) Sparse Carex curvula meadows in the Balkan region (Todorka, Bulgaria). (D) Two contrasting situations: a site with high cover of C. curvula (FAG – Făgăraș Mountains, Southern Carpathians, Romania) and a site with low vegetation cover (MAD – Pic de Madamète, Pyrenees, France).

Fig. 2.

(A) Tussock of Carex curvula from the Balkan region (BEZ – Bezbog). (B) The two anatomical traits used in this study. Three examples are shown to illustrate both (1) the increasing proportion of bulliform cells forming the total leaf cross-sectional area and (2) the increasing circularity of the leaf cross-section. These examples represent samples from the Alps (SIL – Silvretta), Balkan Peninsula mountains (KAB – Kabul) and Pyrenees (MUN – Muntanyó). Population details can be found in Supplementary Data Table S1.

Fig. 3.

(A) Hierarchical clustering of the sites based on the selected combination of traits, namely height, SLA, leaf strength, C:N, bulliform cells and circularity. Three functional groups (F1–F3) were derived from the dendrogram to identify the main trait syndromes. Colour codes as in Fig. 1. (B) Distribution of population means for different traits according to the three functional groups obtained from the hierarchical clustering. Different letters indicate significant differences between populations (post hoc Tukey’s test; P < 0.05). Abbreviations of traits as in Table 1.

Fig. 4.

Coordination among traits assessed by bivariate relationships using standardized major axis (SMA) regressions. Coloured lines indicate significant linear relationships for populations in functional group F1 (blue) or F2 + F3 (red). Grey lines indicate significant relationships for the full dataset. Abbreviations of traits as in Table 1.

Fig. 5.

(A) Distribution of the within-population coefficient of variation (CV) of traits according to the three functional groups obtained from the hierarchical clustering. Abbreviations of traits as in Table 1. (B) Distribution of within-population Nei’s genetic diversity index across the three functional groups. The same letters indicate non-significant differences among groups (post hoc Tukey’s test; P < 0.05).