Adaptability of Swiss Stone Pine (Pinus cembra) in Two Different Environmental Conditions of Romanian Carpathians

Featured Application

Afforestation of mixed spruce–pine stands more resistant to windfall at the upper altitudinal limit of Romanian forests.

Abstract

Swiss stone pine (Pinus cembra) is a feasible solution for increasing the resistance of future mixed spruce–pine stands to windfall at the upper altitudinal limit of Romanian forests. This study aims to analyze the adaptability of ten full-sib families in two different environmental conditions and their evolution in time for predicting forward selection. At a seedling age of 33, evaluations were carried out for survival rate (Sr), growth (diameter at breast height—Dbh, and trees’ height), and branch traits, as well as for stem straightness. The high level of genetic variability, inheritance rate, and trait–trait correlations, registered in both trials, ensure the success of the breeding program. The Swiss stone pine shows a very good adaptation on high-sloping lands from the upper altitudinal limit of Romania’s forests, with the cross-pollinated (CP) families registering a 28.5% better survival in the Rodna trial (34° slope) compared to the Păltiniș experiment (7° slope). The consanguineous families registered only an 11% Sr in Păltiniș, while in the Rodna trial, the Sr was four times higher (in both trials, consanguineous trees were dominated). The juvenile–adult correlations of growth traits were significant, indicating that early selection could be efficient. The forward selection of the best 10% of CP families for Dbh could be applied, ensuring a 9% to 15% genetic gain.

1. Introduction

Because of the high ability of Swiss stone pine (Pinus cembra L.) to adapt to the limiting climatic conditions of high altitude, at the upper limit of the forests, in subalpine areas, many researchers recommend this species [1,2,3,4,5,6,7,8]. Unfortunately, the species faces a reduction in genetic diversity due to isolation or range fragmentation [9,10,11,12,13,14], which requires an increase in the number of seed stands established after a genetic diversity investigation [15,16,17]. Its natural distribution covers the subalpine zones (altitudes between 1200 and 2300 m, a.s.l.) of Europe, the Alps and Carpathian Mountains, in Austria, the Czech Republic, France, Germany, Italy, Poland, Romania, Switzerland, and Ukraine [18,19]. Swiss stone pine presents good resistance to blister rust [20] and also to drought events [21,22].

In Romania, Ioan Blada coordinated a huge breeding program for five-needle pines, including Swiss stone pine, in the 1979–2019 period, resulting in an afforested area of 50 ha in Pietrosul Rodnei mountain (Eastern Carpathians), three seed orchards [23], and also nine field trials (three maternal, two full diallel, three of provenances, and one with clones) [7,24,25,26,27,28,29,30,31].

Swiss stone pine is important because it can be a feasible solution for increasing the resistance of future mixed spruce–pine stands to windfall at the upper altitudinal limit of Romanian forests. For this purpose, the present study aims to analyze the adaptability of ten full-sib families in two different environmental conditions, and their evolution in time in order to predict the optimal age of forward selection. Research on Swiss stone pine is quite limited, especially in recent years, but the species will return to foresters’ attention as a response to the strong and increasingly frequent storms that cause huge damage in pure stands with Norway spruce. Even less in quantity is information from the comparative trials of Swiss stone pine for trees whose age is over 30 years. The working hypotheses of the present research aim to carry out the following: (i) the analysis of genetic variability between and within families and the behavior of CP and SP families in different environmental conditions; (ii) the analysis of correlations between growth and adaptability traits, and juvenile–adult correlations, in order to establish the improvement strategy and the optimal age for forward selection; and (iii) the determination of heritability and genetic gain in order to establish the optimal trait for selection and also the genetic inheritance rate. The scientific questions to which we sought answers were as follows: (a) Are the Swiss stone pine families adapted to the difficult environmental conditions of the upper altitudinal limit of Romania’s forests? (b) What is the optimal age for forward selection, the most indicated trait for selection, and the genetic inheritance rate?

2. Materials and Methods

In 1989, ten Swiss stone pine parental trees were selected in the Southern Carpathians (Retezat mountain, Gemenele population, near the Gemenele lake, Figure 1), at an altitude of 1850 m (a.s.l.), and a 10 × 10 full-diallel mating design was established [26,32]. The ten parental trees were selected according to their phenotypic growth characteristics, to which were added the presence of fruiting and the existence of a distance of at least 50 m among them. Controlled pollination took place in the middle of July 1989, while the seeds were collected in September 1990, and sown in the valley of Bogdan forestry nursery (Prahova County, near Sinaia city) at the end of October, 1990 [29].

In the autumn of 1996, when the seedlings were 6 years old, a comparative trial was established in the Cibinului mountains (Southern Carpathians, near Păltiniș, in Sibiu County, Figure 1), at 1650 m altitude, 45°63′ north latitude, and 23°92′ east longitude (WGS 84 coordinate system). The 100 families (90 from out-crossing and 10 from self-pollination) were planted in a completely randomized block design with four replicates, 15 trees per unitary plot, and 2.5 × 2.5 m spacing [29]. The trial was managed by the forest district Rășinari, and located in production unit V Oncești, plots 89P and 90P, on an area of 3.7 ha. The experiment was situated on a higher, undulating slope, with a southern exposition and a 7° slope inclination. The biotope was represented by a mountain spruce forest of low to medium productivity and the soil was a typical Prepodsoil. The forest type was a normal spruce with Vaccinium myrtillus flora [33]. The average annual temperature was 3.2 °C, and the sum of annual precipitation was 1062 mm [34]. While the precipitation showed no clear trend within the last century, the temperature has increased in the last 20 years (Figure 1).

In the autumn of 1998, when the seedlings grown in the nursery reached the age of 8 years, a second field trial was established in the Rarău mountains (Eastern Carpathians, near Sângeorz-Băi, in Bistrița-Năsăud County, Figure 1), at 1440 m altitude, 47°53′ north latitude, and 24°77′ east longitude (WGS 84 coordinate system). Due to the lack of sufficient seedlings, the experimental device consisted of a 5 × 5 full-diallel cross and a factorial in which the remaining five trees had only the role of fathers. In the two extra years spent in the nursery, the seedlings grew only 5 cm in height and 2 mm in collar diameter. The 25 full-diallel families (20 from out-crossing and 5 from self-pollination) were planted in a completely randomized block design with four replicates, 10 trees per unitary plot, and 3 × 3 m spacing. The 20 factorial families (all from out-crossing) were planted in the same scheme and distances. The trial was managed by the forest district Feldru, located in production unit Anieșu Mare, plot 276, on an area of 1.6 ha. The experiment was situated on a higher, undulating slope, with a northwest exposition and a 34° slope inclination. The biotope is represented by a mountain spruce forest of low to medium productivity and the soil is a typical districambosoil. The forest type is normal spruce with Oxalis acetosella-Dentaria bulbifera flora [35]. The average annual temperature was 3.2 °C (similar with Păltiniș), and the sum of annual precipitation was 1265 mm, 19% higher than in Păltiniș test [34]. While the precipitation showed no clear trend within the last century, the temperatures have increased in the last 20 years (Figure 1).

Figure 1. Location of the Păltiniş and Rodna field trials (red triangles) and provenance origins (red circle) on the Pinus cembra distribution map [36]. Natural species distribution is hatched (A). Regional climate trends (B) in the mean annual temperature (lines) and total precipitation (columns) for the trials during the last 28 years were computed using the climate downscaling tool [34].

Figure 1. Location of the Păltiniş and Rodna field trials (red triangles) and provenance origins (red circle) on the Pinus cembra distribution map [36]. Natural species distribution is hatched (A). Regional climate trends (B) in the mean annual temperature (lines) and total precipitation (columns) for the trials during the last 28 years were computed using the climate downscaling tool [34].

After 26 and 25 growing seasons in trials, in the spring of 2023 (Păltiniș trial, seedlings with 32 growing seasons) and, respectively, in the autumn of 2023 (Rodna trial, seedlings with 33 growing seasons; one year difference in trials growing seasons is compensated by one year difference in seedlings age), measurements and evaluations were carried out for all the existing trees for growth and quality traits as follows:

-

Diameter at breast height (Dbh), using a forest caliper;

-

Tree height (Th) using a Vertex V instrument;

-

Average branch diameter (Bd) in the whorl situated closest to 1.3 m from the ground, measured with electronic calipers;

-

The stem straightness (Ss) was evaluated on a scale of 5 according to Treebreedex protocol [37] as follows: 0 = vertical and rectilinear stem, 1 = 1–2 small curves, 2 = three or more small curves, 3 = 1–2 big curves, and 4 = three or more big curves.

At the office, the branches’ finesse was calculated as follows: Bf = (Bd/Dbh) × 100.

Also, the survival rate (Sr) was calculated at the family level as a percentage between the number of trees that had survived until the age of evaluation and the number of planted trees.

Statistical Analysis

The comparisons of cross-pollinated (CP) with self-cross-pollinated (SP) were made in the Statistica 10.0 [38] and R software [39]. Factorial ANOVA was used for analyzing the influence of environment, genetic components, and the interactions between them, while Duncan and Tukey’s HSD post hoc tests were used to separate the families into homogeneous groups and analyze the significance of the differences among them.

For Păltiniș 10 × 10 diallel mating design, Griffing’s [32] method 3 was used, where one set of F1s and reciprocals but not the consanguineous were included, while for full-sib of Rodna, Griffing’s [32] method 1 was used. The mathematical model for diallel analysis was taken from Wilcox [40], and is a combination of Hayman [41] and Griffing’s [32] models, as follows:

x_kij = u + g_i + g_j + s_ji + m_i + n_ij + e_kij

where x_kij = the value of the kth progeny from the cross between i^th female and j^th male; u = the general mean; g_i and g_j = the general combining ability effects for the i^th and j^th parents, respectively; m_i = the maternal effect of the i^th parent; s_ij = the specific combining ability effect for the cross between the i^th and j^th parents; n_ij = the interaction effect between the i^th and j^th parents, such that r_ij = −r_ji (reciprocal effect not accounted for by maternal or paternal effects); and e_kij = random error [29].

The additive variance was used to estimate the variance of GCA among all of the parents and was estimated as 1/4σ²_A, assuming that all the epistatic genetic variances were irrelevantly small. Also, the estimated variance of SCA was an estimator of 1/4σ²_D [29,42,43].

In both trials, for breeding strategy, two types of heritabilities were calculated, and, consequently, two genetic gains were determined, as follows:

-

Broad-sense family mean heritability (h²₁):

h²₁ = (2σ²_GCA + σ²_SCA)/σ²_Ph = (2σ²_GCA + σ²_SCA)/(2σ²_GCA + σ²_SCA + σ²_Mat + σ²_Mat-Int + σ²_e/k)

where σ²_GCA, σ²_SCA, σ²_Mat, σ²_Mat-Int, σ²_e, and σ²_Ph = general, specific, maternal, maternal interaction, error, and phenotypic variances, respectively, and k = number of blocks = 4.

According to Falconer [44], the genetic gain (G₁) was estimated as follows:

ΔG₁ = i₁ × h²₁ × σ_Ph

where i₁ is the selection intensity proposed by Becker [43] and σ_Ph is the family mean phenotypic standard deviation.

-

Narrow-sense family mean heritability (h²₂):

h²₂ = 2σ²_GCA/σ²_Ph = (2σ²_GCA)/(2σ²_GCA + σ²_SCA + σ²_Mat + σ²_Mat-Int + σ²_e/k)

The genetic gain (G₂) was estimated as follows:

ΔG₂ = i₁ × h²₂ × σ_Ph1

Phenotypic correlations among the analyzed traits, and juvenile–adult correlations in each trial were calculated in R [39].

3. Results

3.1. Phenotypic Variability

For the survival rate (Sr) in more difficult environmental conditions (especially due to very high slope inclination) of the Rodna trial, the average of the CP families was 64%, superior by 28.5% to the Sr of the CP families from the Păltiniș trial (Figure 2). ANOVA revealed significant differences among replications in both trials which highlights the importance of replication trial establishment to reduce the microenvironmental effect. Only 11% of the consanguineous (self-cross-pollinated, SP) tree combinations survived in Păltiniș, while in the Rodna trial, the Sr of the inbred individuals was 48%, but in both trials, most of the SP were dominated and could disappear in the coming years. Among the common five mothers (code 2, 3, 209, X, and Y) the highest Sr was registered for the trees reported to mother 3 (69.6% in Rodna and 52.5% in Păltiniș), while the lowest Sr (59.6% in Rodna and 42.3% in Păltiniș) was observed for mother 209 (Figure 2). A highly significant influence (p < 0.001) of locality was registered by applying factorial ANOVA, while mother and mother x locality interaction has an insignificant influence (

Table 1. Factorial ANOVA for the analyzed traits.

Factor	MSSr	MSDbh	MSTh	MSBd	MSBf	MSSs
Locality	22,181 ***	509 ***	0.73	3.53 ***	1248 ***	135.4 ***
Mother	953	21.95 ***	1.35 ***	0.41 ***	9.84 *	2.57 **
Locality × Mother	252	4.20	0.81 **	0.16	8.01	0.37
Error	413	3.43	0.22	0.08	3.69	0.55

MS_Sr, MS_Dbh, MS_Th, MS_Bd, MS_Bf, and MS_Ss are the mean square for survival rate, diameter at breast height, trees’ height, branches’ diameter, branches’ finesse, and stem straightness. Significant influences: * at p < 0.05, ** at p < 0.01, and *** at p < 0.001 according to the Fisher test.

). However, the mother influence was very close to the significant level (p = 0.058).

For diameter at breast height (Dbh), the CP families registered in the Păltiniș trial had an average value of 16.2 cm, superior by 15% to the Rodna CP mean value. For the common five families, the superiority is even higher, at 17%. Also, for SP, a 10.7% superiority was registered in Păltiniș. The CP tree combinations registered 37% and 42% higher mean values than SP in the Rodna and Păltiniș trials.

The best mother trees (X) in both trials (Figure 3) were superior by 4.3% and 9.3% to the trial averages, and by 8.9% and 13.5% (statistically significant) to the last mother (3, among the common five mothers), in Rodna and Păltiniș, respectively. Highly significant influences (p < 0.001) of locality and mother were registered (Table 1).

For tree height (Th), exactly the same average value, 5.9 m, was registered for CP in both trials, superior by 14% and 34% to SP, in Rodna and Păltiniș, respectively. The average of the common five families in Păltiniș was slightly higher (6.0 m). The best mother trees (X in Păltiniș and 2 in Rodna) were superior by 3.4% and 5.1% to the trial averages, and by 7.0% and 10.7% (both statistically significant) to the last mother (3, among the common five mothers, in Păltiniș and 209 in Rodna). Although it did not rank first, according to Tukey’s HSD post hoc test, mother X is also part of the best homogenous group in the Rodna trial. A highly significant influence (p < 0.001) of mother and a distinctly significant influence (p < 0.01) of locality x mother interaction were registered for Th, the only trait which has not been influenced by locality (Table 1).

Regarding branches (Bd and Bf), the average trial values are 8% and 21.6% lower in Păltiniș compared to Rodna for Bd and Bf, respectively, and the branches of the CP families, although they are thicker with 11.6% (Păltiniș, for the common five mothers) and 25.5% (Rodna) than SP, in regard to their stem, these are 22% (Păltiniș) and 8% (Rodna) finer. A highly significant influence (p < 0.001) of locality and a significant to highly significant influence of mothers were registered by applying factorial ANOVA, while the locality x mother interaction was not significant but with p-value (0.09 for Bd and 0.07 for Bf) close to significant threshold (Table 1). The trees that are related to the most valuable mother in terms of growth (X) are part of the most valuable homogeneous group for Bf in both trials (

Table 2. Duncan test for branches’ diameter (Bd), branches’ finesse (Bf), and stem straightness (Ss) in the Rodna and Păltiniș trials. Homogeneous groups (Hg, the same vertical *) for α = 5%.

Trial	Mother	Bd -cm-	Hg			Mother	Bf -%-	Hg			Mother	Ss -index-	Hg
Rodna	3	2.48	*			2	18.8	*			2	1.82	*
	2	2.56	*	*		3	18.8	*			X	1.95	*
	Y	2.62		*		Y	18.9	*			209	1.99	*	*
	209	2.65		*	*	X	19.1	*			3	2.01		*
	X	2.74			*	209	19.4	*			Y	2.26		*
Păltiniș	Y	2.27	*			Y	14.3	*			2	0.68	*
	2	2.35	*	*		X	14.3	*			X	0.69	*
	3	2.41		*	*	2	14.9	*	*		3	0.71	*	*
	209	2.45			*	209	15.4		*	*	209	0.75		*	*
	X	2.50			*	3	15.6			*	Y	0.96			*

).

Both the CP and SP families, in both trials, registered good results for stem straightness (Ss), that is, a healthy tree, with vertical and rectilinear stems, with 1–2 small curves on some trees. A highly significant influence (p < 0.001) of locality and a distinctly significant (p < 0.01) influence of mother were registered by applying factorial ANOVA (Table 1). The average Ss of the five common mothers were 0.76 and 2.01 in the Păltiniș and Rodna trials, with the best result registered by the 2 and X mothers (Table 2).

3.2. Trait–Trait and Juvenile–Adult Correlations

The growth traits (Dbh and Th) were positively and highly significantly correlated (r = 0.52 *** in Rodna and r = 0.75 *** in Păltiniș), and also significantly and negatively (favorable) correlated with the stem straightness (r = −0.22 *** and −0.26 *** in Rodna and r = −0.04 and −0.13 * in Păltiniș), which allowed forward selection after a growth trait (

Table 3. Trait–trait (up) and juvenile–adult (down) correlations for CP families in Rodna (above diagonal) and Păltiniș (below diagonal).

Trait–Trait	Dbh	Th	Bd	Bf	Ss
Dbh	-	0.52 ***	0.51 ***	−0.56 ***	−0.22 ***
Th	0.75 ***	-	0.15 *	−0.43 ***	−0.26 ***
Bd	0.63 ***	0.37 ***	-	0.42 ***	0.06
Bf	−0.59 ***	−0.61 ***	0.20 ***	-	0.29 ***
Ss	−0.04	−0.13 *	0.04	0.08	-
Juvenile–Adult	Thadult	Thjuv.	Dbhadult	Dbhjuv.
Thadult	-	0.42 *	x	0.44 *
Thjuv.	0.72 ***	-	0.56 **	x
Dbhadult	x	0.50 *	-	0.43 *
Dbhjuv.	0.47 *	x	0.71 ***	-

Dbh, Th, Bd, Bf, and Ss as in Table 1. Th_adult and Dbh_adult are the current values, while Th_juv. and Dbh_juv. are the juvenile values, at the seedling age of 14 years. Significant correlations: * at p < 0.05, ** at p < 0.01, and *** at p < 0.001.

). The correlations of the growth traits with the branch characteristics are unfavorable (r = 0.15 * and 0.51 *** in Rodna and r = 0.37 *** and 0.63 *** in Păltiniș for Bd), which suggests the application of a two-stage breeding strategy for the trees with higher Th and lower Bd inside of the best families (the X one).

For the CP families, significant to highly significant juvenile–adult correlations (Table 3) were registered both for Th (r = 0.42 * in Rodna and r = 0.72 *** in Păltiniș) and Dbh (r = 0.43 * in Rodna and r = 0.71 *** in Păltiniș), which suggests that in the future, forward selection could be practiced even from the juvenile stage, at the seedlings age of 14 years. Both for Th and Dbh, at the adult and juvenile ages, the best performing mother tree was X, with an insignificant exception for Th in the Rodna trial. At the opposite pole, mother 3 registered low results in both trials and at the two analyzed ages. The age–age trends differ significantly for the CP and SP families. Thus, if at the seedling age of 7 years, the average values for Th and Dbh were almost identical, at the adult age, CP was superior by 14–42%.

3.3. Genetic Inheritance and Forward Selection

3.3.1. Heritability

In the Rodna trial, for the analyzed traits, the inheritance rate was medium to high, with the highest heritability being recorded for Bd, followed by Th. The broad-sense family heritability (h²₁) ranged from 0.35 (Bf) to 0.63 (Bd), while the narrow-sense heritability (h²₂) increased from 0.18 (Dbh) to 0.37 (Bd). The Th registered a medium inheritance rate (

Table 4. Estimates of heritabilities (h²) and genetic gains (ΔG%) based on family selection.

Trials	Păltiniș				Rodna
Parameter/Trait	h21	h22	ΔG1	ΔG2	h21	h22	ΔG1	ΔG2
Dbh	0.51	0.35	13.4	9.2	0.47	0.18	15.3	11.8
Th	0.60	0.39	5.4	3.5	0.59	0.33	6.7	3.7
Bd	0.35	0.32	3.9	3.5	0.63	0.37	5.5	3.2
Bf	0.44	0.35	5.4	4.2	0.35	0.22	6.5	3.6

h²₁, h²₂, ΔG₁, and ΔG₂ are the broad-sense and narrow-sense family heritability, and the specific genetic gains (for selection of the best 10% CP combinations). Dbh, Th, Bd, and Bf as in Table 1.

).

In the Păltiniș experiment, the inheritance rate was medium to high, with the highest heritability being recorded for Th, followed by Dbh. The broad-sense family heritability (h²₁) ranged from 0.35 (Bd) to 0.60 (Th), while the narrow-sense heritability (h²₂) increased from 0.32 (Bd) to 0.39 (Th). The Dbh and Bf registered a medium inheritance rate (Table 4).

These results indicate a medium to high level of heritability, which allows us to carry out the forward selection strategy of Swiss stone pine for growth and branch traits.

3.3.2. Genetic Gain and Forward Selection

The genetic gain that could be achieved was estimated at the family level (Table 4) by selecting the best 10% of the families (combinations reported in the majority to mother X). In the Rodna trial, if the best 10% of the families are selected and multiplicated vegetatively, a genetic gain (ΔG₁) of 5.5% (Bd) to 15.3% (Dbh) will be achieved. At the same intensities, if the multiplication takes place sexually, the genetic gains (ΔG₂) range between 3.2% for Bd and 11.8% for Dbh (Table 4).

In the Păltiniș experiment, if the best 10% of the families are selected and multiplicated vegetatively, a genetic gain (ΔG₁) of 3.9% (Bd) to 13.4% (Dbh) will be achieved. At the same intensities, if the multiplication takes place sexually, the genetic gains (ΔG₂) range between 3.5% for Th and Bd and 9.2% for Dbh (Table 4).

Therefore, the best genetic gain will be obtained by the selection of the best 10% of the families (majority with X mother) for Dbh (a situation in which Th will also be improved based on high correlations with Dbh in both trials).

4. Discussion

4.1. Phenotypic Variability

A high variability among and within the families for all the traits was recorded, which favors the forward selection. These results are in accordance with the earlier findings [29], and available only for the Păltiniș trial. This fact is hopeful if we refer to the molecular genetic investigations made in the Alps and the Carpathians, which showed a low genetic diversity and an amplified risk of inbreeding [10,11,12]. At the same time, other investigations reported a higher differentiation and genetic diversity in Swiss stone pine populations from the Carpathians compared to the Alps’ ones [45,46].

In the present research, the comparisons of the mean performances of CP and SP for the growth (Th and Dbh) traits showed highly significant differences in favor of CP (14–42%), again, in agreement with the earlier data [29]. Only 11% in Păltiniș and 48% in Rodna of the consanguineous trees have survived, and present an underdeveloped state of vegetation and a major risk of disappearance in the coming years. Similar results were reported in the Romanian Carpathians for Norway spruce [47].

4.2. Trait–Trait and Juvenile–Adult Correlations

Significant and positive correlations among the growth traits were registered in accordance with previous results, reported for the most important European tree species, Norway spruce [48] and European beech [49], as well as for the species studied in the present research, in juvenile stage [7,29].

The juvenile–adult growth traits correlations were moderate to high and statistically significant, demonstrating that early selection might be efficient. Because of that, in the future, selection could be practiced even from the juvenile stage, at the trial age of 14 years. Earlier, Blada and Popescu [29] suggested an even lower age of 6 years for growth trait selection.

4.3. Genetic Inheritance and Forward Selection

The high level of genetic variability registered for all of the analyzed traits, as well as the high number of families involved in these experiments, ensure the success of the breeding strategy.

Both in the Rodna and Păltiniş field trials, the narrow-sense were smaller than broad-sense heritabilities for all the growth traits, in accordance with the previous findings [29]. For the growth traits, after an increased trend in heritability in the first 10 years after planting, in the last twenty years, the values were maintained. An increasing trend in heritability in the first 8 years for growth traits was also found for Pinus taeda [50], while for Pinus contorta, an opposite trend in the first years after planting was registered [51]. Also, a constant trend in the time of family heritability for growth traits was observed for Pinus pinaster [52].

In the Păltiniș trial, the genetic gain increased in the last nineteen years for Dbh. Consequently, if the most valuable families were selected, the broad-sense genetic gains increased from 6% at 7 years after plantation time [29] to 13.4% at present. For Th, the broad-sense genetic gain was reduced significantly in the same period, from 11% [29] to 5.4% at present.

Consequently, based on the results obtained in the present paper, for both trials, we propose the forward selection of the best 10% of the CP families for Dbh, which ensures an important genetic gain (9–15%). The selected combinations of the most valuable 10% families (most of them reported to mother X) will be mixed with other clones from different trials in order to also ensure a high genetic diversity in future trials, seed orchards, or afforested lands.

5. Conclusions

The high level of genetic variability, inheritance rate, and trait–trait correlations, registered for all of the analyzed traits (especially for growth traits) in both trials, as well as the high number of families involved in this experiment (especially in the Păltiniș trial), ensure the success of the breeding program.

The Swiss stone pine shows a very good adaptation to high-sloping lands from the upper altitudinal limit of Romania’s forests, registering a 28.5% better survival in the Rodna trial (an average slope of 34°) compared to the Păltiniș experiment (an average slope of 7°). The consanguineous families registered only an 11% Sr in Păltiniș, while in the Rodna trial, the Sr of inbred was 48%, but in both trials, most of the consanguineous trees were dominated and could disappear in the coming years.

The juvenile–adult correlations of growth traits were significant, indicating that early selection, even from the juvenile stage (at the age of 14 years), could be efficient. Based on the medium to high heritabilities and the estimated genetic gain, the forward selection of the best 10% of the CP families (especially of those with mother X) for Dbh could be applied, which ensures at least a 9.2% genetic gain.

Future research directions should focus on genomic and transcriptomic analyses, the investigation of the species’ resilience to climate warming, and assisted migration evaluation in order to expand the knowledge about Pinus cembra, which is particularly important in Romania because it can mix with Picea abies in the upper altitudinal limits of forests.