Phenotypic variability, inheritance rate and selection strategy of Pinus cembra half-sib families in two common garden experiments in the Southern Carpathians

Schlüsselbegriffe: Alpenraum, Zuchtstrategien, Feldversuche, quantitative Genetik, Subalpine Wälder, Zirbe

Available at https://doi.org/10.53203/fs.2503.1

Download as PDF

See below the issue 3/2025 as E-Paper or have a look at our E-Paper archive dating back to 1955.

Abstract

In the interest of facilitating the reintroduction of Swiss stone pine (Pinus cembra L.) in the Lotrului and Șureanu mountains (Southern Carpathians), where the species naturally existed in the past, two half-sib comparative trials were conducted at Cugir and Voineasa. In total, 71 and respectively 65 families were tested (41 common in both trials), which belong to 7 provenances (4 common) originating from two countries (Romania and France) and three mountain divisions (French Alps, Eastern and Southern Carpathians). This study aims to analyze genetic variability by examining testing site influences, inheritance rates, and trait correlations in two 27-year-old half-sib trials with the goal of identifying the best-performing families and developing an optimal breeding strategy to enhance desired characteristics in future generations At the trial’s 27 years of age, measurements and evaluations of the breast height diameter, tree height, branch diameter and finess of branches, stem straightness, and survival were performed, while the trait-trait, trait-origin, and inheritance rate were estimated. The high level of genetic variance and heritability (especially for tree height), along with the significant correlations among traits, and the large number of families, ensures the success of the breeding program. The MGIDI index (multi-trait genotype-ideotype distance) allowed us to select ten to eleven families in each trial that could be promoted in the Southern Carpathians and under similar ecological conditions to those in the trial where they were performed. A forward selection of the top 30% of individual trees from the best selected families, based on tree height, could be applied in each trial. Given the strong effects of the testing site and site x family interaction, extreme attention is required when transferring reproductive materials. Our findings provide a constructive approach for management plans, as we recommend transition from Norway spruce monocultures to structurally more diverse stands with codominant Swiss stone pine, at the upper altitudinal limit of the Southern Carpathian forests. 

Zusammenfassung

Zur Wiederansiedlung der Zirbelkiefer (Pinus cembra L.) in den Lotrului- und Șureanu-Gebirgen (Südkarpaten), wo die Art in der Vergangenheit natürlich vorkam, wurden in zwei Halbsippen-Vergleichsversuchen – Cugir und Voineasa – 71 bzw. 65 Familien getestet (davon 41 in beiden Versuchen gemeinsam). Diese Familien stammen aus sieben Herkünften (vier davon gemeinsam), die aus zwei Ländern (Rumänien und Frankreich) und drei Gebirgsregionen (Französische Alpen, Ost- und Südkarpaten) stammen. Ziel der Studie ist es, die genetische Variabilität, den Einfluss des Prüfstand­orts, die Vererbungsraten sowie die Korrelationen zwischen Merkmalen und zwischen Merkmalen und Herkünften zu analysieren, um die leistungsfähigsten Familien auszuwählen und die geeignetste Züchtungsstrategie vorzuschlagen. Nach 27 Jahren Versuchsdauer wurden der Brusthöhendurchmesser, die Baumhöhe, die Aststärke und -feinheit, die Geradschaftigkeit sowie das Überleben der Bäume gemessen und bewertet. Gleichzeitig wurden die Merkmalskorrelationen, Herkunftsabhängigkeiten und Erblichkeiten geschätzt. Der hohe Grad an genetischer Varianz und Heritabilität (insbesondere bei der Baumhöhe), die signifikanten Merkmalskorrelationen sowie die große Zahl getesteter Familien bilden eine solide Grundlage für ein erfolgreiches Züchtungsprogramm. Der Einsatz des MGIDI-Index (multi-trait genotype-ideotype distance) ermöglichte die Auswahl von zehn bis elf Familien je Versuch, die für eine Förderung in den Südkarpaten und unter ähnlichen ökologischen Bedingungen wie am Prüfstandort geeignet sind. Bei jedem Versuch konnte eine Vorwärtsauswahl der besten 30 % der Einzelbäume aus den am besten ausgewählten Familien auf Grundlage der Baumhöhe vorgenommen werden. Aufgrund der starken Einflüsse des Prüfstandorts sowie der Standort-Familien-Interaktion ist jedoch bei der Übertragung von Vermehrungsmaterialien größte Sorgfalt geboten. Unsere Ergebnisse liefern einen konstruktiven Ansatz für Managementpläne, da wir den Übergang von Fichtenmonokulturen zu strukturell vielfältigeren Beständen mit kodominanter Zirbelkiefer an der oberen Höhengrenze der Wälder der Südkarpaten empfehlen.

1 Introduction

Subalpine forests of the Alps and Carpathians represent highly sensitive ecosystems (Körner, 2012) which have experienced a more significant temperature change than the global average (Jochner, 2017), a trend expected to persist, with both potential pressures and opportunities for tree species (Ulber et al., 2004; Tranquillini, 2012; Ponocná et al., 2016; Jandl et al., 2019; Dauphin et al., 2021; Obojes et al., 2024; Marčiš et al., 2025). In this situation, it is necessary to identify species and intraspecific varieties that show adaptability to climate change. The genetic plasticity, gene flow, comigration and demographic dynamics of these species/ varieties could reinforce resilience by accelerating the adaptive reactions (Kremer et al., 2025).

Swiss stone pine (Pinus cembra L.) is an important tree species at the alpine treeline of the European Alps and Carpathians (Tóth et al., 2019; Gugerli et al., 2022). The species' weak apical dominance enables flexible growth responses to environmental stresses, such as frost and snow, thereby enhancing its survival in challenging alpine conditions (Oberhuber et al., 2019). Nevertheless, a strong reaction of Swiss stone pine to climate limiting factors, was registered (Carrer et al., 2007). The low temperature during the growing season (< 5ºC) is a limiting factor (Körner, 2012; Etzold et al., 2021), but the most restrictive issue seems to be the water regime, especially the drought (Casalegno et al., 2010; Unterholzner et al., 2024). However, Swiss stone pine exhibits a remarkable potential for adaptation to drought events (Știrbu et al., 2022; Izworska et al., 2023; Muter et al., 2024), which is superior to that of other species, such as Norway spruce (Popa et al., 2017). A careful consideration of genetic diversity, assisted gene flow, environmental factors, and seasonal dynamics is required to ensure the species' adaptability and the long-term viability (Farjon, 2013; Mosca et al., 2016; Spathelf et al., 2018; Ammer, 2019; Bastin et al., 2020; Fragnière et al., 2022; Budeanu et al., 2024, 2025a,b; Chakraborty et al., 2024; Sonnenwyl et al., 2024).

In the past, Swiss stone pine had a much broader distribution across Europe and Siberia. After the last ice age and during warmer periods, its range became fragmented, causing it to retreat to higher elevations in the Alps and Carpathians. Researchers have identified Pinus cembra as a species of five-needle pine in Europe and Pinus cembra sibirica as a subspecies distributed especially in Siberia, the differentiation is likely caused by gradual adaptation. A new relative to Swiss stone pine, Pinus pumila, appears in tundra climate, and extends from Siberia to North Korea and Japan (European Wilderness Society, 2023; Gugerli et al., 2023).

In Romania, a breeding programme for promoting Swiss stone pine was coordinated by Blada (1987, 1999, 2019), which have involved the establishment of nine field trials: three half-sib, two full-sib, three of provenances, and one of clones (Blada & Popescu, 2007, 2008, 2012; Budeanu et al., 2024, 2025a, b). The importance of the species, in Romania and in Europe, is mainly due to its ability to adapt to the difficult conditions at the upper altitudinal limit of the forests, but the high-quality wood with industrial use (the best among pines), the edible seeds with pharmaceutical uses, and the pleasant ornamental appearance should not be neglected (Șofletea & Curtu, 2007). The Swiss stone pine could be an alternative three species for the future in order to replace the spruce monocultures (affected by windfalls and snowbreaks) with species mixtures. Previous research has highlighted very significant growth and adaptation differences among the different provenances tested in the Romanian Carpathians (Blada & Popescu, 2007, 2008, 2012), which confirms the importance of conducting comparative studies.

This study aims to analyze genetic variability by examining the effects of testing sites, inheritance rates, and trait correlations in two 27-year-old half-sib trials, with the goal of identifying the best-adapted families and developing an optimal breeding strategy to enhance desirable traits in future generations. For the two half-sib trials that have reached the age of 27 years, nothing has been published until this article, even though they are part of the breeding program mentioned above, and can contribute with valuable families and populations to ensure a high level of genetic diversity in the future afforestation.

2 Materials and methods

In the interest of facilitating the reintroduction of Swiss stone pine in the Lotrului and Șureanu mountains (Southern Carpathians), where the species naturally existed in the past, two half-sib comparative trials were conducted at Cugir and Voineasa. In total, 71 families were tested in the Cugir trial, while 65 families were tested in the Voineasa trial, with 41 families being common to both trials. These families belong to 7 provenances (4 common) originated from two countries (Romania and France) and from three mountain divisions (French Alps, Eastern and Southern Carpathians). The place of origin (including altitude) of the seven provenances is shown in Figure 2, while a more detailed presentation can be found in the previous articles (Blada & Popescu, 2007; Budeanu et al., 2025a, b). Significant differences in latitudinal (4°27’), longitudinal (20°29’), and altitudinal (695 m) amplitudes exist among the provenances (Blada & Popescu, 2007; Figure 2). In each of the seven provenances, seeds were harvested from 4 to 20 open-pollinated trees, located at least 50 m apart (Blada & Popescu, 2007), and the seeds were kept separately per tree, so that each family represents the descendants of a single tree.

The experiments were established in the autumn of 1997, using 7-year-old seedlings, planted at a distance of 2.5 x 2.5 m, in a completely randomized block design, with 71 families x 3 replications x 10 seedlings/unitary plot, in Cugir trial, and respectively 65 x 3 x 10, in Voineasa test.

The Cugir half-sib trial is located in the Șureanu Mountains, and it is currently administered by the Cugir Forest District (Alba County), being located in the production unit IV Canciu, plot 65B, covering an area of 1.3 ha, near to the provenance trial analyzed in Budeanu et al. (2025a,b). Geographically, it is located at 45˚37’ N latitude, 23˚32’ E longitude, and at an average altitude of 1500 m. The trial is situated at the middle of a slope on land with an undulating configuration, facing southwest, with a 23˚ inclination. The biotope is of medium productivity, with the forest type being represented by high-altitude spruces with mull flora, and the soil is typically dystric Cambisol.

The Voineasa half-sib trial was installed in the Lotrului Mountains of the Southern Carpathians, and it is administered by the Voineasa Forest District. It is situated in the production unit VII Haneș, plot 141B, covering an area of 1.5 ha (including an important area with reserves). Geographically, it is located at 45˚30’ N latitude, 23˚47’ E longitude, and at an average altitude of 1650 m. The trial is situated at the top of a slope on land with an undulating configuration, facing north-east, with a 23˚ inclination. The biotope is of medium productivity, with the forest type being represented by high-altitude spruces with mull flora, and the soil is classified as typically dystric Cambisol.

According to the European forest type classification (2007), the both trials are included in the Subalpine and mountainous spruce forests, into the Subalpine - mountainous Carpathian mountains forest type (Ozenda, 1988). The mull flora is represented especially by Oxalis acetosella, Asperula odorata, Dentaria bulbifera, and Vaccinium myrtillus (Beldie & Chiriță 1967).

In both testing regions, while the precipitation showed no clear trend over the last 30 years, temperatures have exponentially increased (from 2.6°C and 1.9°C, in 1996-1999, period to 4.2°C and 3.9°C, in 2020-2024 period, in Cugir and Voineasa respectively). The average annual temperature of the last three decades were 3.8°C in Cugir, and 3.0°C in Voineasa, while the sum of annual precipitation averaged 1080 mm in Cugir and 1190 mm in Voineasa (B4EST, 2025).

At the trial’s 27 years of age (34 years tree age), measurements and evaluations of growth and quality traits were carried out for all existing trees. The breast height diameter (Dbh) measured at 1.3 m from the ground, tree height (Th), and dominant branch diameter (Bd, measured at the whorl situated at 1.3 m from the ground) were assessed using a forest caliper, a Vertex V instrument, and an electronic caliper, respectively. The stem straightness (Ss) was evaluated on a scale of 5 classes (0 = vertical and rectilinear stem, 1 = 1-2 slight curves, 2 = three or more slight curves, 3 = 1-2 big curves, and 4 = three or more big curves), according to Ducci et al. (2012). Additionally, the finesse of branches (Bf) was calculated as Bf=Bd x 100 / Dbh. Survival (S) was calculated in both trials and for each family as the proportion of surviving trees from the total number of planted trees.

 

2.1 Statistical analysis

In each trial, the mathematical model used for analyses of the variance (ANOVA) was the one recommended by Nanson (2004), with the replication (environmental factor, integrated in the model as a fixed effect), provenance and family, as genetic factors (as random effect in the mathematical model). After that, the results for common families were cumulated and analyzed using a bifactorial analysis to highlight the influences of the testing site (locality, integrated as a fixed effect), family, and their interactions (integrated in the model as a random effect), according to Nanson (2004). Before applying ANOVA, the Kolmogorov-Smirnov test was applied to check the normal distribution of variables, and the assumptions of ANOVA were verified using Levene’s test. The significance level was determined using Fisher (F) test, assessing the transgression probability at 5%, 1% and 0.1% levels. Provenances and families ranking, the differences among them, and their separation into homogenous groups were conducted using the Tukey HSD test at a 5% transgression possibility. Additionally, Pearson’s simple phenotypic correlations were determined, along with correlations between traits and the ecological gradients of the provenance’s origins. The ecophysiological latitude (Le) was estimated according to Wiersma (1962): Le = L + A / 100, where L represents the north latitude and A represents the altitude (a.s.l.).

A graphical method was used to highlight the best-performing provenances, based on the average values recorded for survival (stability performance) and trees’ height (growth performance), using the R programming language (R Core Team, 2025). The metan package (Olivoto & Lúcio, 2020) was used to perform a multi-trait genotype-ideotype distance analysis (MGIDI index) for selecting the best-performing families based on all the analyzed traits (Th, Dbh, S, Bd, Bf, Ss) and to calculate the genetic parameters, including genotypic variances, broad-sense and mean-based heritabilities, as well as the genotypic coefficient of variation.

The software utilized for data processing includes R 4.5.0 (2025), for figures 3-6 and tables 2-4, Statistica 10.0 (2010), for Tukey HSD test (table 1), and QGIS 3.44.1 Solothurn (2024), for Figure 2.

3 Results

3.1 Survival and phenotypic variability

For Survival (S), a high average value was recorded in the Cugir trial (87%), significantly superior to the average of the Voineasa test (69.2%). No significant differences among the five provenances were registered in the Cugir trial, with the highest result (90.3%) obtained by the Alps provenance (Bois de Ayes, France). In the Voineasa common garden, two groups of provenances were separated, with the fourth common provenances (Gemenele, Călimani, Stâna de Râu, and Pietrele) included in the best group, while the provenances Păpușa-Iezer and Valea Lalei being part of the second group (Figure 3). At the family level, a high homogeneity has resulted in the Cugir test, with 70 of the 71 families included in the best group, and nine families (four from France, three from SC, and two from EC) achieving maximum survival (100%). In Voineasa’s experiment, the best result (96.7%) was obtained by a family originating from Valea Lalei, although the average of that provenance was the lowest in the trial (Figure 3). Nineteen of the sixty-five families tested in the Voineasa trial registered a survival of more than 80%, indicating that the selection can be successfully applied.

In Cugir trial, analyses of the variance (ANOVA, Table S1) show a highly significant influence (p < 0.001) of the family factor, and a significant (p < 0.05) impact of the provenance factor, while the microenvironmental factor (replication) played a non-significant role (p > 0.05) for survival. In the common garden of Voineasa, all three factors played a highly significant role. Factorial ANOVA highlighted a highly significant influence of locality, while family had a non-significant role (Table S1).

In both field trials, for the principal growth traits (Th and Dbh), the Pietrele provenance registered the highest results, clearly distinguished from all others. In contrast, among the four common provenances, Gemenele exhibited consistently low performance, particularly in terms of tree height (Figure 4). An identical Th average values were registered in both trials (6.8 m), but with a higher variability (expressed by standard deviation, and by 13, compared to 5, homogeneous groups) in the Cugir trial. For Dbh, a 4% higher average value was registered in the Cugir common garden experiment, with the best provenance, Pietrele, getting almost the same result in both trials (Figure 4). Some provenances exhibited a strong reaction to the changing environment, with good results in Cugir and low in Voineasa (Stâna de Râu provenance) and vice versa (Călimani provenance). At the family level, the Tukey HSD test emphasizes a higher variability for Th than Dbh, in both trials, but especially in the Cugir test, and different families obtained good results (top 10%) in the two trials, highlighting the need for caution when making recommendations regarding specific provenances for afforestation. ANOVA and factorial ANOVA support this conclusion, as the majority of genetic and environmental factors showed a significant influence (p < 0.05) on the growth traits (Table S1).

The provenance highlighted for growth in both trials, Pietrele (SC), also exhibited high survival, ranking as the best homogeneous group in both experiments, and being the first candidate for selection.

Regarding branches (Bd, Bf), the provenances highlighted for growth and survival, Pietrele (SC), in both trials, seconded by Stâna de Râu (SC) and Bois de Ayes (France Alps), in the Cugir test, and by Călimani (EC), in the Voineasa test, also presented thin and fine branches, which are indicative of superior stem wood quality (Figure 4, Table 1). In general, no significant influences of the environmental factor (replication) on the branches’ traits were registered in both trials (except for Bd, in Voineasa), while the genetic factors (family and provenance) played a significant role (Table S1). Factorial ANOVA revealed that all the genetic and environmental factors had a highly significant influence (p < 0.001) on the finesse of branches, while for Bd, the influence of family and Locality x family interaction was insignificant (Table S1).

Stem straightness (Ss) is an important trait for high-quality wood production. Likewise, for the branches traits and in contrast to the growth traits, the average result recorded in the Voineasa trial (0.38) was clearly more favorable compared to the Cugir test (0.85). Thus, a total number of 49 families are included in the zero class (perfectly straight stem, Ss < 0.5) in the test located at a higher altitude, compared to only five families in the Cugir experiment. However, good results for this trait were observed in both trials, with all of the families classified in the 0-1 classes (with a maximum of 1-2 slight stem curves). For Ss, the provenance from the Franch Alps (Bois de Ayes) was the best in the Cugir test, followed by Pietrele (SC). In the Voineasa trial, a high homogeneity was registered with all the provenances included in one homogeneous group (Table 1). ANOVA and factorial ANOVA revealed that all the genetic and environmental factors had a significant (p < 0.05) to highly significant influence (p < 0.001) on the stem straightness (Table S1).

3.2 Trait-trait and trait-origin correlations

Highly significant phenotypic correlations were registered among the principal growth traits (Th and Dbh), and in both trials (r= 0.72*** - 0.74***), while both were also direct and significant correlated with Bd, but with a low intensity for Th (r= 0.15* in Cugir, and 0.39*** in Voineasa). The survival was also positively correlated with all growth traits, including Bd (Table 2). The trait-trait correlations in the two trials suggest that the selection for Th also favors the improvement of most other growth and stability traits, with the smallest negative impact on Bd.

Correlations between traits and the geographical gradients of the provenance’s origin (Table 3), generally indicated a low and statistically insignificant influence of provenance origin in the Cugir trial, while a positive and highly significant influence of the altitude was registered in the Voineasa test, located at a higher altitude. The highest correlation was found between Dbh and altitude (r= 0.39***), suggesting that the provenances from higher altitudes registered superior growth in the Voineasa experiment. Altitude had a negative and significant influence on stem and branches traits (r= -0.15* to r= -0.19**), in the Cugir trial, suggesting that provenances from higher altitudes registered favorable qualitative traits. An opposite result was registered in the Voineasa trial, for Bd (r= 0.31***). Longitude showed a positive and distinctly significant influence on stem and branches traits (r= 0.18** - 0.19**), in the Cugir trial, suggesting that the families from the west (France Alps) presented better stem and branches traits. An opposite effect of longitude on all traits was observed in the Voineasa trial, where French families were not represented.. Another interesting and significant correlation, observed in both trials, was between latitude and survival (Table 3), indicating that provenances from northern regions presented a superior stability.

3.3 Inheritance rate and selection strategy

The percentage of total phenotypic variance explained by genotype, Gen (%), registered the highest values for Th, 25.3% in Cugir, and 37.5% in Voineasa, generating also the highest broad-sense (h²) and mean-base (h²mg) family heritabilities (Table 4). Superior variability for Th in the Cugir test, expressed by ANOVA (Table S1) and Tukey tests, is due to microenvironment and within-family interactions, since the genetic variance has a greater weight in the Voineasa test. Also, the highest accuracy of genotypic value prediction (0.71 in Cugir and 0.80 in Voineasa, closest to the ideal value of 1), and the stronger genetic signal [CV ratio, expressed by ratio of genotypic coefficient of variation to residual (environmental) coefficient of variation; 0.58 in Cugir, and 0.77 in Voineasa trials, again the closest to the ideal value of 1], were obtained for Th (Table 4).

In the Cugir trial, located in the Șureanu mountains of the Southern Carpathians, the mean performance of provenances for the most important growth (Th) and stability (S) traits indicated that the best-performing provenances were Bois de Ayes (from the Franch Alps), Stâna de Râu and Pietrele (both from the Southern Carpathians). Due to the high survival of all provenances, the most important is Pietrele, which registered the highest Th (Figure 5). In the Voineasa experiment, located also in the Southern Carpathians but at a higher altitude, the best-performing provenance was Pietrele, followed by Călimani (from Eastern Carpathians), Gemenele and Stâna de Râu. In both trials, the local provenances, Stâna de Râu and especially Pietrele, registered very high Th values and high Sr values (> 70%), being the most important for the selection strategy (Figure 5).

At the family level, in the Cugir common garden experiment, the multi-trait genotype-ideotype distance index (MGIDI) allows us to select the best eleven most performing families (Figure 6) based on all the analyzed traits (Th, Dbh, S, Bd, Bf, Ss). The best families belong mainly to Bois de Ayes (3) and Pietrele (6), while two other provenances, Călimani and Stâna de Râu, are each represented by only one family. In the Voineasa trial, MGIDI index indicates the selection of the top ten best-performing families (Figure 6), which are part of Călimani (4), Pietrele (1), Stâna de Râu (2), Gemenele (1), and Păpușa-Iezer (1) provenances. Consequently, the best inheritance rate could be obtained by a mixed selection of the best 30% of trees from the ten (Voineasa) and eleven (Cugir) most valuable families for Th, which could ensure improvements in all growth and stem quality traits.

4 Discussion

4.1 Phenotypic variation

In the present paper, the last in a series of five articles concerning the variability of Swiss stone pine conserved ex situ in common garden experiments, we analyzed the 65 to 71 families originated from seven provenances (one from Alps, two from Eastern Carpathians, and four from Southern Carpathians), in two different environments of the Southern Carpathians (different mountains, 150 m difference in altitude, opposite exposition, sunny vs. shady, slope position, etc.). These differences contributed to enhanced adaptability in the more favorable conditions of the Cugir test (survival was superior with 18%, compared to the Voineasa trial). Almost the same survival was registered in the European provenances trial Cugir (established beside of the family experiment, on top of it), while in the paired trial (Cârlibaba, located in the Eastern Carpathians), situated at 100 m lower altitude, a slightly superior survival was registered (Budeanu et al., 2025b), which could suggest a good adaptation at lower altitude, and a feasible solution for mixture with Norway spruce. The growth traits’ results, especially for Dbh, were also superior in the Cugir trial, again in accordance with the other trials’ results (Budeanu et al., 2025b), while the qualitative traits of the stem and branches were higher in the Voineasa experiment. The genetic factor seems to have a significant role, as a provenance Pietrele achieved good performances in both experiments (actually, in all six testing sites, Budeanu et al., 2024, 2025a,b), but the necessity for attention when transferring the reproductive materials is emphasized by the opposing results recorded in the two trials by provenances Călimani and Stâna de Râu. A similar concern was underlined in the previous papers (Budeanu et al., 2024, 2025a). ANOVA confirms the strong influence of the microenvironment, in each trial, and a highly significant influence of the testing site conditions. The high variability, among and within families, is encouraging for the success of selection, while the adapted families could play an essential role in forest resilience to global warming (Chakraborty et al., 2024), which will demand constant adaptation of the tree species (Leonelli et al., 2009; Mosca et al., 2016; Neuschulz et al., 2018; Gugerli et al., 2022).

 

4.2 Trait-trait and environmental Correlations

In Cugir and Voineasa half-sib trials, the phenotypic correlations favor the selection for growth traits (Dbh or Th), which also ensures the improvement of the majority of stem and branches qualitative traits (except of Bd). Similar results were previously registered for the same species (Budeanu et al., 2024, 2025a,b). The positive correlations between growth traits and the altitude of the family’s origin, registered in both trials but highly significant only in Voineasa’s higher altitude test, indicate a strong influence of altitude on the Swiss stone pine distribution, results similar to the research performed in the Tatra Mountains (Zięba et al., 2019; Izworska et al., 2023). It appears that populations from higher altitudes exhibit superior adaptive traits, more pronounced in the test at the highest altitude (Voineasa). Previously, in experiments located at slightly lower altitudes, such correlations were not significant (Budeanu et al., 2024, 2025a,b). The global warming may limit the Swiss stone pine distribution at lower altitudes, in accordance with the findings from the Western Carpathians (Marčiš et al., 2025).

 

4.3 Inheritance rate and forward selection

In both experiments, the highest heritabilities were registered for Th, the most important growth trait for this hard growing species, so, the breeding program should prioritize the selection based on Th, which will ensure the competitiveness of Swiss stone pine with the others tree species of the high altitude phytoclimatic area, especially the Norway spruce and European larch.

For the joint results of both trials, the best-performing provenance in terms of stability and growth was Pietrele, which originates from the same region where the trials were established. However, the selection only in favor of local provenances is not justified, since in the Cugir and Voineasa trials, the second Th result was obtained by the Bois de Ayes provenance (from the French Alps, in Cugir), and the Călimani provenance (from Eastern Carpathians, in Voineasa). Due to the highly significant influence of the testing site, we must restrict forward selection to the provenance region of the testing sites (C1, the northern part of the Southern Carpathians) and recommend the best families of the above-mentioned provenances for use in similar environmental conditions. To increase the genetic diversity, all 21 families highlighted by MGIDI could be selected, with special attention to Th. Thus, we propose the selection of the best 30% of trees from the most valuable 21 families for Th. Moreover, to increase the genetic diversity in future afforestation, it is necessary to combine seedlings from these common gardens with those highlighted in the other four trials (Budeanu et al., 2024, 2025a,b). Lendvay et al. (2014) underline the importance of a sufficiently large population size to guarantee a high level of genetic diversity, which is a critical problem for Pinus cembra in Europe (Belokon et al., 2005; Höhn et al., 2009; Salzer & Gugerli, 2012; Dzialuk et al., 2014; Wojnicka-Półtorak et al., 2015).

Our findings provide a constructive approach for management plans, as we recommend transition from Norway spruce monocultures to structurally more diverse stands with codominant Swiss stone pine, at the upper altitudinal limit of the Southern Carpathian forests. The conservation plans in favor of Swiss stone pine must include the preservation of the natural forests, of all trials and seed orchards, and ensure the extension of the species in the spruce phytoclimatic floor of the Southern Carpathian (Budeanu et al., 2025b). The establishment of a monitoring network at the upper altitudinal limit of the forests would be useful, as would be the highlight of the economic value of Swiss stone pine timber (Koeck et al. 2025).

5 Conclusions

The high level of genetic variances and heritabilities (especially for tree’ height), along with the significant correlations among traits, coupled with the large number of families involved in this experiment, firmly establishes the foundation for a successful breeding program.

Following 27 years of testing, ten to eleven families of Swiss stone pine were successfully identified in each trial, reflecting a selection intensity of 15%. These families are now considered the most suitable candidates for promotion in the northern region of the Southern Carpathians, where the ecological conditions closely resemble those of the experimental sites. Furthermore, a forward selection of the top 30% of individual trees from the best selected families, based on tree height, could be applied in each trial. Given the strong effects of the testing site and site x family interaction, extreme attention is required when transferring reproductive materials.

Authors' contributions

Conceptualization, M.B. and E.B.; methodology, M.B. and E.B.; formal analysis, M.B. and E.B.; investigation, M.B. and E.B.; writing-original draft preparation, M.B.; writing-review and editing, E.B.; visualization, M.B. and E.B. All authors have read and agreed to the published version of the manuscript.

Funding and Acknowledgments

This paper was financed by the Romanian UEFISCDI, in the frame of 71PHE-Carmine project contracted by National Institute for Research and Development in Forestry “Marin Drăcea”. We wish to express our gratitude to Gabriela Grosu and Dan Pepelea for their help with field measurements.

Declarations

The authors declare no competing interests. The authors declare that they have no known conflicts of interest.

References

Ammer C (2019) Diversity and forest productivity in a changing climate. New Phytol 22: 50–66. https://doi.org/10.1111/nph.15263

Bastin JF, de Haulleville T, Maniatis D, Marchi G, Massaccesi E, Mollicone D, Papa C, Pregagno C (2020) Tree restoration potential in the European Union. FAO and European Commission Directorate General for Environment (DG ENV). https://doi.org/10.13140/RG.2.2.24811.67368/1

B4EST (2025) Adaptive breeding for better forests. Climate database available online: http:// www.b4est.eu/ (accessed on March, 2025).

Belokon MM, Belokon YS, Politov DV, Altukhov YP (2005) Allozyme polymorphism of Swiss stone pine Pinus cembra L. in mountain populations of the Alps and the Eastern Carpathians. Russ J Genet 41: 1268–1280. https://doi.org/10.1007/s11177-005-0228-0

Beldie A, Chiriță C (1967) Indicative flora of our forests (Flora indicatoare din pădurile noaste). Agrosilvică Publishing House, Bucharest, 216 p. (In Romanian)  

Blada I (1987) Cercetări asupra rezistenței genetice la Cronartium ribicola a unor hibrizi interspecifici (Research on the genetic resistance to Cronartium ribicola of some interspecific hybrids). PhD thesis, A.S.A.S., Bucharest, 146 p. (in Romanian)

Blada I (1999) Diallel crossing in Pinus cembra. III. Analysis of genetic variation at the nursery stage. Silvae Genet 48(3-4): 179–187.

Blada I, Popescu F (2007) Swiss stone pine provenance experiment in Romania. II Variation in growth and branching traits to age 14. Silvae Genet 56(3-4): 148–158.

Blada I, Popescu F (2008) Diallel crossing in Pinus cembra. IV. Age trends in genetic parameters and genetic gain for growth and branching traits. Ann For Res 51: 89–113.

Blada I, Popescu F (2012) Diallel crossing in Pinus cembra: age trends in genetic parameters and genetic gain for height. Silvae Genet 61(1–2): 66–79. https://doi.org/10.1515/sg-2012-0009

Blada I (2019) Memoriu de activitate 1961-2007 (Activity memory 1961–2007). Silvică Publishing House, Bucharest, 51 p. ISBN: 978-606-8020-62-4 (in Romanian)

Budeanu M, Popescu F, Besliu E, Apostol EN (2024) Adaptability of Swiss stone pine (Pinus cembra) in two different environmental conditions of Romanian Carpathians. Appl Sci 14(16), 7428, 12 p. https://doi.org/10.3390/app14167428

Budeanu M, Besliu E, Pepelea D (2025a) Testing the radial increment and climate–growth relationship between Swiss stone pine European provenances in the Romanian Carpathians. Forests 16(3), 391, 15 p. https://doi.org/10.3390/f16030391

Budeanu M, Popescu F, Apostol EN, Pleșca IM, Besliu E (2025b) Adaptability of twelve European provenances of Pinus cembra in two different branches of the Carpathians. Silvae Genet 74(1): 63-76. https://doi.org/10.2478/sg-2025-0007

Carrer M, Nola P, Eduard JL, Motta R, Urbinati C (2007) Regional variability of climate–growth relationships in Pinus cembra high elevation forests in the Alps. J Ecol 95: 1072–1083. https://doi.org/10.1111/j.1365-2745.2007.01281.x

Casalegno S, Amatulli G, Camia A, Nelson A, Pekkarinen A (2010) Vulnerability of Pinus cembra L. in the Alps and the Carpathian Mountains under present and future climates. For Ecol Manag 259(4): 750–761. https://doi.org/10.1016/j.foreco.2009.10.001

Caudullo G, Welk E, San-Miguel-Ayanz J (2017) Chorological maps for the main European woody species. Data in Brief 12: 662–666. https://doi.org/10.1016/j.dib.2017.05.007

Chakraborty D, Ciceu A, Ballian D, et al. (2024) Assisted tree migration can preserve the European forest carbon sink under climate change. Nat Clim Change 14: 845–852. https://doi.org/10.1038/s41558-024-02080-5

Dauphin B, Rellstab C, Schmid M, Zoller S, Karger DN, Brodbeck S, Guillaume F, Gugerli F (2021) Genomic vulnerability to rapid climate warming in a tree species with a long generation time. Glob Change Biol 27: 1181–1195. https://doi.org/10.1111/gcb.15469

Ducci F, De Cuyper B, Proietti R, Pâques L, Wolf H (2012) TREEBREEDEX, research infrastructure network 2006-2011. Reference protocols for assessment of traits and reference genotypes to be used as standards in international research projects. CRA SEL Publishing House, Arezzo, Italy, 82 p.

Dzialuk A, Chybicki I, Gout R, Maczka T, Fleischer P, Konrad H, Curtu AL, Şofletea N, Valadon A (2014) No reduction in genetic diversity of Swiss stone pine (Pinus cembra L.) in Tatra Mountains despite high fragmentation and small population size. Conserv Genet 15: 1433–1445. https://doi.org/10.1007/s10592-014-0628-6

Etzold S, Sterck F, Bose AK, Braun S, Buchmann N, Eugster W, Gessler A, Kahmen A, Peters RL, Vitasse Y, Walthert L, Ziemińska K, Zweifel R (2021) Number of growth days and not length of the growth period determines radial stem growth of temperate trees. Ecol Lett 25: 427–439. https://doi.org/10.1111/ele.13933

European Wilderness Society (2023). Unknow story of Pinus cembra. Available at: https://wilderness-society.org/unknow-story-of-pinus-cembra/#:~:text=Research%20proofed%20that%20in%20Alps,sibirica%20in%20Siberia.

European Environment Agency (2007) European forest types: categories and types for sustainable forest management reporting and policy. 2nd edition, Copenhagen, Denmark, 111 p. Available at: http://www.env-edu.gr/Documents/European%20forest%20types.pdf

Farjon A (2013) IUCN Red List of Threatened Species: Pinus cembra [WWW Document]. IUCN Red List Threat. Species. Available at: https://www.iucnredlist.org/en

Fragnière Y, Sonnenwyl V, Clément B, Kozlowski G (92022) Large-scale historical afforestation failure with Pinus cembra in the Swiss Prealps. New For 53: 533–553. https://doi.org/10.1007/s11056-021-09871-0

Gugerli F, Brodbeck S, Bebi P, Bollmann K, Dauphin B, Gossner M, Krumm F, Peter M, Queloz V, Reiss G, Rellstab C, Stofer S, von Arx G, Wasem U, Zweifel R (2022) Swiss stone pine - portrait of a mountain forest tree. Fact sheet 72, 16 p. https://doi.org/10.55419/wsl:32467

Gugerli F, Brodbeck S, Lendvay B, Dauphin B, Bagnoli F, van der Knaap WO, Tinner W, Höhn M, Vendramin GG, Morales-Molino C, Schwörer C (2023) A range-wide postglacial history of Swiss stone pine based on molecular markers and palaeoecological evidence. J Biogeogr 50: 1049–1062. https://doi.org/10.1111/jbi.14586

Höhn M, Gugerli F, Abran P, Bisztray G, Buonamici A, Cseke K, Hufnagel L, Sebastiani F, Quintela-Sabaris S, Vendramin GG (2009) Variation in the chloroplast DNA of Swiss stone pine (Pinus cembra L) reflects contrasting postglacial history of populations from the Carpathians and the Alps J Biogeogr 36: 1798–1806. https://doi.org/10.1111/j.1365-2699.2009.02122.x

Izworska K, Muter E, Matulewski P, Zielonka T (2023) Tree rings as an ecological indicator of the reaction of Swiss stone pine (Pinus cembra L.) to climate change and disturbance regime in the extreme environment of cliff forests. Ecol Indic 148, 110102, 12 p. https://doi.org/10.1016/j.ecolind.2023.110102.

Jandl R, Spathelf P, Bolte A, Prescott CE (2019) Forest adaptation to climate change-is non-management an option? Ann For Sci 76, 48, 13 p. https://doi.org/10.1007/s13595-019-0827-x

Jochner MA (2017) Treelines in the Swiss Alps: growth dynamics and forest succession in a changing climate (PhD Thesis). ETH Zurich, Switzerland, https://doi.org/10.3929/ethz-b-000264159

Koeck R, Kessler M, Dorfstetter Y, Oberklammer I, Klosterhuber R, Starlinger F, Lick H, Englisch M, Schaufler J, Keßler D, Lexer MJ, Vacik H (2025) Altitudinal distribution of Stone Pine (Pinus cembra L.) in the Styrian Alps as indication for silvicultural stability and resilience strategies. Austrian J For Sci 142: 113-136.

Körner C (2012) Alpine treelines: functional ecology of the global high elevation tree limits. Springer Basel, Switzerland, https://doi.org/10.1007/978-3-0348-0396-0

Kremer A, Chen J, Lascoux M (2025) ‘Chimes of resilience’: what makes forest trees genetically resilient? New Phytol 246: 1934–1951. https://doi.org/10.1111/nph.70108

Lendvay B, Höhn M, Brodbeck S, Mîndrescu M, Gugerli F (2014) Genetic structure in Pinus cembra from the Carpathian Mountains inferred from nuclear and chloroplast microsatellites confirms post-glacial range contraction and identifies introduced individuals. Tree Genet Genomes 10: 1419–1433. https://doi.org/10.1007/s11295-014-0770-9

Leonelli G, Pelfini M, Battipaglia G, Cherubini P (2009) Site-aspect influence on climate sensitivity over time of a high-altitude Pinus cembra tree-ring network. Clim Change 96: 185–201. https://doi.org/10.1007/s10584-009-9574-6

Marčiš P, Rybár J, Šebeň V, Murgaš V, Sitková Z (2025) Picea abies and Pinus cembra at high altitudes show different growth reaction to rising temperatures: Study from the Western Carpathian subalpine forests. Dendrochronologia 91(4), 126325. https://doi.org/10.1016/j.dendro.2025.126325

Mosca E, Gugerli F, Eckert AJ, Neale DB (2016) Signatures of natural selection on Pinus cembra and P. mugo along elevational gradients in the Alps. Tree Genet Genomes 12, 6. https://doi.org/10.1007/s11295-015-0964-9

Muter E, Izworska K, Wilczyński S, Zielonka T (2024) Climate change differentially alters the climate-growth relationships of Norway spruce and Swiss stone pine in the Western Carpathians. Preprint. https://doi.org/10.2139/ssrn.4947559

Nanson A (2004) Genetic and forest trees breeding. Ed. Gembloux, Belgium, 712 p. ISBN: 978-287-0160-70-1 (in French)

Neuschulz EL, Merges D, Bollmann K, Gugerli F, Böhning-Gaese K (2018) Biotic interactions and seed deposition rather than abiotic factors determine recruitment at elevational range limits of an alpine tree. J Ecol 106: 948–959. https://doi.org/10.1111/1365-2745.12818

Oberhuber W, Geisler TA, Bernich F, Wieser G (2019) Weak apical control of Swiss stone pine (Pinus cembra L.) may serve as a protection against environmental stress above treeline in the central European Alps. Forests 10(9), 744. https://doi.org/10.3390/f10090744

Obojes N, Buscarini S, Meurer AK, Tasser E, Oberhuber W, Mayr S, Tappeiner U (2024) Tree growth at the limits: the response of multiple conifers to opposing climatic constraints along an elevational gradient in the Alps. Front For Glob Change 7. https://doi.org/10.3389/ffgc.2024.1332941

Olivoto T, Lúcio AD (2020) Metan: An R package for multi‐environment trial analysis. Methods in Ecol Evol 11: 783-789. https://doi.org/10.1111/2041-210X.13384. 

Ozenda P (1988) The vegetation of the Alps in the European mountain region (Die Vegetation der Alpen im Europäischen Gebirgsraum). Stuttgart‑New York: Gustav Fischer Verlag Publisher, 353 p. (in German)

Ponocná T, Spyt B, Kaczka R, Büntgen U, Treml V (2016) Growth trends and climate responses of Norway spruce along elevational gradients in East-Central Europe. Trees 30: 1633–1646. https://doi.org/10.1007/S00468-016-1396-3

Popa I, Nechita C, Hofgaard A (2017) Stand structure, recruitment and growth dynamics in mixed subalpine spruce and Swiss stone pine forests in the Eastern Carpathians. Sci Total Environ 598: 1050–1057. https://doi.org/10.1016/j.scitotenv.2017.04.169

R Core Team (2025) R: A language and environment for statistical computing; R foundation for statistical computing: Vienna, Austria. Available online: https://www.r-project.org/ (accessed on April, 2025).

QGIS development team (2024) QGIS Geographic Information System. Open-source geospatial foundation project. Available online: http://qgis.osgeo.org (accessed on March 2025)

Salzer K, Gugerli F (2012) Reduced fitness at early life stages in peripheral versus core populations of Swiss stone pine (Pinus cembra) is not reflected by levels of inbreeding in seed families. Alp Bot 122: 75–85. https://doi.org/10.1007/s00035-012-0106-z

Sevdari K, Marmullaku D (2023) Shapefile of European countries. Technical University of Denmark. Dataset. https://doi.org/10.11583/DTU.23686383

Sonnenwyl V, Dauphin B, Fragnière Y, Clément B, Grünig S, Brodbeck S, Parisod C, Kozlowski G, Gugerli F (2024) Genetic underpinning of historical afforestation with allochthonous Pinus cembra in the northwestern Swiss Alps. Alp Bot 134: 1–13. https://doi.org/10.1007/s00035-023-00304-6

Spathelf P, Stanturf J, Kleine M, Jandl R, Chiatante D, Bolte A (2018) Adaptive measures: integrating adaptive forest management and forest landscape restoration. Ann For Sci 75, 55, 6p. https://doi.org/10.1007/s13595-018-0736-4

Statistica 10.0 (2010) StatSoft Inc., Tulsa, OK, USA.

Șofletea N, Curtu L (2007). Dendrology (Dendrologie). Transilvania University Publishing House, Brașov, Romania, 418 p. (in Romanian)

Știrbu M-I, Roibu C-C, Carrer M, Mursa A, Unterholzner L, Prendin AL (2022) Contrasting Climate Sensitivity of Pinus cembra Tree-Ring Traits in the Carpathians. Front Plant Sci 13. https://doi.org/10.3389/fpls.2022.855003

Tóth EG, Tremblay F, Housset JM, Bergeron Y, Carcaillet C (2019) Geographic isolation and climatic variability contribute to genetic differentiation in fragmented populations of the long-lived subalpine conifer Pinus cembra L. in the western Alps. BMC Evol Biol 19, 190, 17 p. https://doi.org/10.1186/s12862-019-1510-4

Tranquillini W (2012) Physiological Ecology of the Alpine timberline: tree existence at high altitudes with special reference to the European Alps. Ecological Studies 31, Springer-Verlag, Berlin, Heidelberg, https://doi.org/10.1007/978-3-642-67107-4

Ulber M, Gugerli F, Bozic G (2004) EUFORGEN. Technical guidelines for genetic conservation and use for Swiss stone pine (Pinus cembra). Biodiversity International. Available at: https://hdl.handle.net/10568/104342

Unterholzner L, Castagneri D, Cerrato R, Știrbu M-I, Roibu C-C, Carrer M (2024) Climate response of a glacial relict conifer across its distribution range is invariant in space but not in time. Sci Total Environ 906, 167512. https://doi.org/10.1016/j.scitotenv.2023.167512

Wiersma JH (1962) Enkete kwantitatieve aspecten van het exotenvraagstuk. Nederlands Bosbouw Tijtschrift 34(5): 175–184.

Wojnicka-Półtorak A, Celiński K, Chudzińska E, Prus-Głowacki W, Niemtur S (2015) Genetic resources of Pinus cembra L. Marginal populations from the Tatra mountains: implications for conservation. Biochem Genet 5: 49–61. https://doi.org/10.1007/s10528-015-9670-4

Zięba A, Różański W, Bukowski M, Ciesielska B, Szwagrzyk J (2019) Distribution and habitat conditions of Pinus cembra forests in the Tatra Mountains. Dendrobiology 81: 86–96. http://dx.doi.org/10.12657/denbio.081.010