Genetic Parameters and Family Selection of Pinus pseudostrobus var. apulcensis Through Growth and Stem Quality in Mixteca Oaxaqueña Region, Mexico

Abstract

The Mixteca Oaxaqueña region has historically suffered from soil and vegetation loss. However, since the last decade of the previous century, successful reforestation efforts have been carried out in many areas. As a result, there is now a need to select phenotypes with superior growth and good wood quality. This study aimed to estimate genetic parameters and identify superior families of Apulco pine (Pinus pseudostrobus var. apulcensis (Lindl.) Shaw) based on growth and stem quality traits. After four years, growth and stem quality traits were evaluated in 64 open-pollinated families. Different selection intensities were tested using two evaluation methods. All traits showed significant differences among families (p ≤ 0.0016), with genetic coefficients of variation ranging from 8.01% to 18.84%. Heritability estimates for growth traits were high ($h_i^2$ = 0.42−0.66; $h_f^2$ = 0.55−0.63), whereas heritability for stem quality traits was slightly lower ($h_i^2$ = 0.01−0.38; $h_f^2$ = 0.01−0.38). Genetic correlations (r_g) among growth traits were high and positive (r_g ≥ 0.857), while correlations among stem quality traits and between quality and growth traits were more variable (r₉ = −0.498 to 0.899). Based on both evaluation methods and a 25% selection intensity, sixteen superior families were identified, showing estimated genetic gains of approximately 4% for growth-related traits. These families are recommended for use in timber plantations, whereas a broader set is suggested for reforestation efforts in order to maintain genetic diversity.

1. Introduction

The genus Pinus L. plays a fundamental role in the forest ecosystems of the Americas due to its ecological significance and economic and social importance [1]. In Mexico, pines represent the most valuable natural resource, owing to their traditional and commercial uses, cultural relevance, and the ecosystem services they provide. They hold high economic value, as they are sources of timber, firewood, pulp, resin, edible seeds, and other products [2,3]. Mexico harbors a high diversity of pine species, with 49 taxa distributed across approximately 75% of the country’s mountainous regions [4]. However, domestic timber production is insufficient to meet demand and is largely offset by imports from other countries [5].

In response to the growing need to meet the demand for high-quality forest products, forest tree breeding programs have gained strategic importance. These programs aim to select and propagate superior genetic materials to maximize desirable traits such as growth, wood quality, and resistance to biotic and abiotic factors [6,7]. In this context, the study of genetic control and genetic correlations among traits is essential for designing efficient selection strategies [6,7].

Native pine species in Mexico are used for reforestation, soil conservation, timber production, and carbon sequestration [2]. Among them, Apulco pine (Pinus pseudostrobus var. apulcensis (Lindl.) Shaw) is the most commonly used for plantation establishment due to its rapid growth and adaptability [2,8], and the quality and workability of its wood, which is used for lumber, pulp, and bioenergy production [9]. This pine variety is most abundant in the mountain ranges of Oaxaca, in southern Mexico [10]; however, it is also widely distributed across mid- to high-elevation areas of the Sierra Madre Occidental, Transverse Volcanic Belt, Sierra Madre del Sur, and the mountain ranges of Central America [10].

Several studies have shown considerable intraspecific genetic variability among Mexican pines, which enables a favorable response to selection in forest tree breeding programs [11,12]. However, in Mexico, forest genetic improvement efforts have focused on a limited number of species [13,14,15], while Apulco pine has been relatively understudied, particularly regarding stem quality, branch number and thickness, and branch insertion angle—traits that are critical for the timber industry. Nevertheless, this pine variety demonstrates good adaptability to complex edaphic conditions, such as those found in the Mixteca Oaxaqueña, a region historically affected by deforestation and soil erosion [16,17].

In the Mixteca Oaxaqueña region, to restore tree vegetation, successful plantations of native and exotic species have been established since the last decade of the twentieth century [16,17,18]. However, landowners require high-quality timber harvests; thus, toward the end of the previous decade, research initiatives began to select genotypes with higher productivity and superior wood quality, including progeny trials of Apulco pine [14,19]. Given the urgent need to identify quality genotypes for establishing timber plantations in the Mixteca Oaxaqueña, early selection of Apulco pine based on genetic variation and control of growth and stem quality traits is essential. Early selection is feasible because the age–age genetic correlation is high for growth and stem quality variables in pines, including smooth-bark Mexican (Pinus pseudostrobus Lindl.) [20,21,22].

Therefore, the objectives of this study were (1) to evaluate differences in growth and stem quality traits among progenies of Apulco pine established in the Mixteca Oaxaqueña region; (2) determine the genetic control of growth and stem quality traits, and the genetic correlations among these traits; and (3) select the best-performing families based on growth and stem quality traits using two evaluation methods. The hypotheses tested were as follows: (1) growth and stem quality traits differ among families due to genetic variation; (2) growth and stem quality traits are under strong genetic control, and their genetic correlations are statistically significant; and (3) both selection methods will identify the same top-performing families and yield similar genetic gains for each trait.

The identification of Apulco pine families with high productive potential and superior stem quality represents a key step towards establishing local breeding programs adapted to the specific conditions of the Mixteca Oaxaqueña. The results of this study are relevant not only for the improvement of this pine variety but also serve as a reference for other understudied forest species. Furthermore, they constitute a valuable contribution by providing empirical evidence of the potential of family selection to enhance productivity and the quality of forest resources.

2. Materials and Methods

2.1. Tree Selection and Establishment of the Progeny Test

In July 2017, a total of 64 Apulco pine (Pinus pseudostrobus var. apulcensis (Lindl.) Shaw) trees were selected in Mexico: 42 from the Sierra Juárez region (Oaxaca), 19 from the Sierra Madre de Chiapas (Chiapas), and 3 from the Sierra Madre Oriental (Veracruz) (Table S1, Figure 1). Selection was based on superior growth and stem quality for timber production [6]. Seeds were collected between November 2017 and February 2018. Subsequently, from March to October 2018, seedlings were produced at the forest nursery of the National Institute of Forestry, Agriculture and Livestock Research (INIFAP), located at 17.188580° N, −96.801210° W, at an elevation of 1621 m a.s.l. Seedlings were cultivated in 310 cm³ rigid containers filled with a substrate composed of 50% peat moss (Flora–gard^®; Floragard Vertriebs GmbH, Oldenburg, Lower Saxony, Germany), 25% perlite, and 25% vermiculite. The substrate was supplemented with Osmocote^® Plus (6.6% N–NO₃, 8.4% N–NH4, 9% P₂O₅, 12% K₂O) (Scotts MiracleGro, Marysville, OH, USA) at a rate of 5 kg per m³ of substrate. During nursery cultivation, foliar fertilization was applied once a week using Folim^® (40% P₂O₅, 40% K₂O; 0.16 g L⁻¹ of water) (Grupo Irbaquim SA de CV, Mexico City, Mexico) and Aminofit^® (13.38% K₂O, 5.31% Ca–EDTA, 3.26% NO₃⁻, 1.43% H₃BO₃; 2 mL L⁻¹ of water) (PT Ultraquimia, Jiutepec, Morelos, Mexico).

In October 2018, a progeny test was established using 64 open-pollinated families of Apulco pine in the community of Cabecera de Cañada, municipality of Santiago Yosondúa, in the Mixteca Oaxaqueña region (16°52.344′ N, 97°03.520′ W), at an elevation of 2385 m, with an average annual temperature of 14.9 °C and annual rainfall of 1108 mm (Figure 1). The soil is classified as clay loam (33.6% sand, 31.6% silt, and 34.8% clay), with 2.09% organic matter and a pH of 6.03 [19]. The concentrations of N, P, K, and Ca are 18.4, 2.26, 208, and 1667.5 mg kg⁻¹, respectively [19]. A randomized complete block design (RCBD) was used, with 20 replications per family and one tree per experimental unit, using a spacing of 3 × 3 m between plants (1280 trees in total). To minimize edge effects, a border row of trees of the same variety was planted around the experimental plot.

2.2. Traits Evaluated

Basal diameter and total height were measured in November 2018 (first measurement) and again in January 2022 (second measurement). Periodic increments in basal diameter and height were calculated by subtracting the 2018 value from the corresponding 2022 value.

In addition, in 2022, crown diameter was measured using a tape measure, along with stem quality traits including stem straightness, branch diameter, number of whorls and branches, and branch insertion angle (Table S2). Stem straightness was evaluated qualitatively on a four-level scale: very crooked (1), crooked (2), slightly crooked (3), and straight (4). The number of branches and whorls was determined by direct counting. Branch diameter and insertion angle were measured using a digital caliper and a protractor, respectively, on two branches (the thickest and its opposite) located 1.0 m above ground level [14].

2.3. Statistical Analysis

Analysis of variance was conducted using the MIXED procedure in SAS statistical software, version 9.4 [23]. Variance and covariance components for all variables were estimated using the VARCOMP procedure with the restricted maximum likelihood (REML) method, which is appropriate when the data are unbalanced [24]. The model used for the randomized complete block design was as follows:

$$Y_{ij}=\mu +B_i+F_j+{\epsilon }_{ij}$$

(1)

where Y_ij= value of the observation of the i-th family in the j-th block (repetition), µ = population mean, B_i = fixed effect of the i-th block, F_j = random effect of the j-th family, ε_jk = error of the effects, i = 1, 2, …, 20 blocks, and j = 1, 2, …, 64 families.

Individual heritability ($h_i^2$) and family mean heritability ($h_f^2$) were calculated using the following equations [14,25]:

$$h_i^2={\displaystyle \frac{3{\sigma }_f^2}{{\sigma }_f^2+{\sigma }_e^2}}$$

(2)

$$h_f^2={\displaystyle \frac{\left(\frac{3}{4}{\sigma }_f^2\right)}{\left({\sigma }_f^2+{\sigma }_e^2/b\right)}}$$

(3)

where ${\sigma }_f^2$ = variance of families, ${\sigma }_e^2$ = the error variance, and b = harmonic mean of the number of plants per family.

To avoid overestimating heritability, additive genetic variance (${{\sigma }_A^2=3\sigma }_f^2$) was calculated using a coefficient of genetic determination of 3, as the families originated from open pollination and likely consist of a mixture of half and full siblings [14,26]. The standard error of individual heritability [SE($h_i^2$)] and the coefficient of genetic variation (CV_g) were calculated using the following equations [14,25]:

$$SE\left(h_i^2\right)={\left[{\displaystyle \frac{{\left(2\left(1+\left(n_f-1\right)\right)h_i^2\right)}^2}{\left(n_a\left(n_a-1\right)\left(n_f-1\right)\right)}}\right]}^{0.5}$$

(4)

$${CV}_g=\left({\displaystyle \frac{{\sigma }_A}{\overline{X}}}\right)\left(100\right)$$

(5)

where n_f = number of families, n_a = number of trees per family, ${\sigma }_A$ = square root of additive variance, and $\overline{X}$ = overall mean.

The phenotypic correlation coefficients between pairs of variables were calculated using Pearson’s correlation. In contrast, genetic correlation coefficients were estimated using the following equation [14,25,26,27]:

$$r_{g\left(XY\right)}={\displaystyle \frac{{COV}_{f(X,Y)}}{\sqrt{({\sigma }_{f\left(X\right)}^2\ast {\sigma }_{f\left(Y\right)}^2)}}}$$

(6)

where ${\sigma }_{f\left(X\right)}^2$ y ${\sigma }_{f\left(Y\right)}^2$ are the variances of families X and Y, respectively; ${COV}_{f(X,Y)}$ is the covariance of families of these variables, obtained by the following equation [28]:

$${COV}_{X,Y}={\displaystyle \frac{{\sigma }_{f(X+Y)}^2-\left({\sigma }_{f\left(X\right)}^2+{\sigma }_{f\left(Y\right)}^2\right)}{2}}$$

(7)

where ${\sigma }_{f\left(X+Y\right)}^2$ is the covariance of families of the variable X + Y. The standard error of the genetic correlation was obtained with the following equation [14,25]:

$${EE(r}_g)=(1-r_g^2)\sqrt{{\displaystyle \frac{EE\left(h_{iX}^2\right)EE\left(h_{iY}^2\right)}{2h_{iX}^2{h}_{iY}^2}}}$$

(8)

Family selection was carried out using two methods: (1) principal component analysis (PCA) [29,30,31] and (2) multi-trait comprehensive evaluation [31,32]. The principal component method was applied using the following equations:

$$V_{ik}=\sum _{j=1}^n\left(U_{jk}\ast Z_{ij}\right)$$

(9)

$$Y_i=\sum _{k=1}^m\left(W_k\ast V_{ik}\right)$$

(10)

where Y_i is the principal component value for family i, Wₖ is the normalized weight of the extracted component k, Vᵢₖ is the score of family i for component k, Uⱼₖ is the loading of trait j on component k, Zᵢⱼ is the standardized value of family i for trait j, m is the number of extracted components, and n is the number of traits.

The multi-trait comprehensive evaluation method was performed using the following equations:

$$a_{ij}={\displaystyle \frac{X_{ij}}{X_{jmax}}}$$

(11)

$$Q_i={\left[\sum _{i=1}^n{a}_{ij}\right]}^{0.5}$$

(12)

where Qᵢ is the comprehensive evaluation score of family i, aᵢⱼ is the evaluation value of family i for trait j, Xᵢⱼ is the mean value of family i for trait j, Xⱼₘₐₓ is the maximum mean value across all families for trait j, and n is the number of traits.

For each evaluation method, genetic gain and realistic gain were obtained for selection intensities of 6.25, 12.5, and 25%, using the following equations [31,33,34]:

$${∆G}_g={\displaystyle \frac{S}{\overline{X}}}\ast h_f^2\times 100$$

(13)

$${∆G}_r={\displaystyle \frac{S}{\overline{X}}}\times 100$$

(14)

where ΔG_g is the genetic gain of the trait, ΔG_r is the realistic gain of the trait, S is the selection differential of the trait, $h_f^2$ is the family mean heritability of the trait, and $\overline{X}$ is the phenotypic mean of the trait among all families.

3. Results

3.1. Differences Between Families and Variance Components

All growth and stem quality traits showed significant differences among families (p ≤ 0.0016) (

Table 1. Significance, overall average, minimum and maximum averages by family, and variance components of growth and stem quality traits in a progeny trial of Apulco pine (Pinus pseudostrobus var. apulcensis) in the Mixteca Oaxaqueña, Mexico.

Traits	p-Value 1	Mean	Standard Error	Average per Family		Variance Component (%)
				Minimum	Maximum	Block	Family	Error
PIBD [cm]	<0.0001	5.67	1.45	3.17	6.68	26.92	11.49	61.59
PIH [m]	<0.0001	1.63	0.50	0.96	2.11	19.16	11.34	69.50
CD [m]	<0.0001	1.36	0.35	0.83	1.73	23.14	17.04	59.82
SS	0.0016	2.99	0.81	2.41	3.60	8.99	2.94	88.07
NW	<0.0001	4.76	1.14	3.25	5.80	1.58	10.35	88.07
NB	<0.0001	18.95	5.37	11.19	25.87	8.52	11.52	79.96
BD [cm]	<0.0001	1.88	0.47	1.37	2.17	14.12	7.86	78.02
BIA [°]	<0.0001	57.09	11.68	42.50	71.55	2.31	11.66	86.03

PIBD, periodic increment in basal diameter; PIH, periodic increment in height; CD, crown diameter; SS, stem straightness; NW, number of whorls; NB, number of branches; BD, branch diameter; BIA, branch insertion angle. ¹ The complete results of the analysis of variance are presented in Table S3.

). The mean values for periodic increment in basal diameter, periodic increment in height, and crown diameter were 5.67 cm, 1.63 m, and 1.36 m, respectively. Stem straightness was classified as slightly twisted. Substantial variation among families was observed for both growth and stem quality traits (Table 1).

For growth traits, the family effect accounted for 11.34 to 17.04% of the total variance, the block effect contributed between 19.16 and 26.92%, and the residual error explained 59.82 to 69.50%. For stem quality traits, the family effect contributed between 2.94 and 11.66% and the block effect between 1.58 and 14.12%, while the residual error had the largest contribution, ranging from 78.02 to 88.07% (Table 1).

3.2. Genetic Variation and Heritability

The coefficients of additive genetic variation for growth and stem quality traits were moderate, ranging from 12.14% to 18.84%, except for stem straightness, which showed lower genetic variation (8.01%). Individual heritability estimates ranged from 0.42 to 0.66 for growth traits and from 0.10 to 0.38 for stem quality traits. In contrast, the heritability of family means ranged from 0.55 to 0.63 for growth traits and from 0.29 to 0.55 for stem quality traits (

Table 2. Coefficient of genetic variation (CV_g), individual heritability ($h_i^2$), heritability of family means ($h_f^2$), and standard error of individual heritability (EE$h_i^2$) in progenies of Apulco Pine (Pinus pseudostrobus var. apulcensis) evaluated in the Mixteca Oaxaqueña.

Variable	CVg (%)	${\mathit{h}}_{\mathit{i}}^2$	EEh2i	${\mathit{h}}_{\mathit{f}}^2$
Periodic increment in basal diameter [cm]	15.11	0.47	0.15	0.57
Periodic increment in height [m]	18.21	0.42	0.15	0.55
Crown diameter [m]	18.84	0.66	0.13	0.63
Stem straightness	8.01	0.10	0.06	0.29
Number of whorls	13.29	0.31	0.13	0.51
Number of branches	16.70	0.38	0.14	0.55
Branch diameter [cm]	12.33	0.28	0.12	0.47
Branch insertion angle [°]	12.14	0.36	0.14	0.51

).

3.3. Genetic and Phenotypic Correlations

Genetic correlations among growth traits were positive and high (r_g = 0.783 to 0.881). Similarly, genetic correlations among stem quality traits were positive and ranged from moderate to high (r_g = 0.335 to 0.899), except for branch diameter, which showed negative correlations with all other stem quality traits (r_g = −0.112 to −0.410). Genetic correlations between growth and stem quality traits were generally positive (r_g = 0.16 to 0.82), except for the correlations of periodic increment in basal diameter with branch diameter and branch angle, and crown diameter with stem straightness, which were negative (r_g = −0.026 to −0.498) (

Table 3. Genetic correlations (above diagonal, standard errors in parentheses) and phenotypic correlations (below diagonal, p-value in parentheses) between growth and stem quality traits in a progeny trial of Apulco pine (Pinus pseudostrobus var. apulcensis) in the Mixteca Oaxaqueña region.

	PIH	PIDB	CD	SS	NW	NB	DB	BIA
PIH		0.857 *	0.738 *	0.514 ns	0.524 *	0.820 *	0.516 ns	0.390 ns
		(0.053)	(0.065)	(0.555)	(0.249)	(0.082)	(0.276)	(0.288)
PIDB	0.838		0.881 *	0.665 ns	0.337 ns	0.553 *	−0.498 ns	−0.245 ns
	(<0.0001)		(0.026)	(0.731)	(0.312)	(0.182)	(0.769)	(0.528)
CD	0.733	0.848		−0.026 ns	0.201 ns	0.376 *	0.561 *	0.160 ns
	(<0.0001)	(<0.0001)		(0.702)	(0.251)	(0.169)	(0.150)	(0.237)
SS	0.519	0.364	0.117		0.899 *	0.873 *	−0.384 ns	0.335 ns
	(<0.0001)	(0.003)	(0.359)		(0.147)	(0.159)	(2.183)	(0.868)
NW	0.460	0.373	0.242	0.585		0.873 *	−0.198 ns	0.468 ns
	(0.0001)	(0.002)	(0.054)	(<0.0001)		(0.073)	(0.867)	(0.318)
NB	0.617	0.542	0.403	0.591	0.814		−0.410 ns	0.461 ns
	(<0.0001)	(<0.0001)	(0.001)	(<0.0001)	(<0.0001)		(0.882)	(0.279)
BD	0.468	0.658	0.666	0.028	0.072	0.100		−0.112 ns
	(<0.0001)	(<0.0001)	(<0.0001)	(0.829)	(0.570)	(0.432)		(0.725)
BIA	0.452	0.326	0.363	0.366	0.529	0.609	0.038
	(0.0002)	(0.009)	(0.003)	(0.003)	(<0.0001)	(<0.0001)	(0.763)

PIBD, periodic increment in basal diameter; PIH, periodic increment in height; CD, crown diameter; SS, stem straightness; NW, number of whorls; NB, number of branches; BD, branch diameter; BIA, branch insertion angle. * p ≤ 0.0353; ^ns p ≥ 0.0615.

).

Phenotypic correlations among growth traits were positive (r = 0.117 to 0.848) and statistically significant (p < 0.0001). Similarly, phenotypic correlations among stem quality traits were positive (r = 0.028 to 0.814) and significant (p ≤ 0.003), except for branch diameter, which showed no significant correlation with any other stem quality trait (p ≥ 0.432) (Table 3). Phenotypic correlations between growth and stem quality traits were also positive (r = 0.117 to 0.666) and significant (p ≤ 0.009), with the exception of crown diameter, which showed no significant correlation with stem straightness and number of whorls (p ≥ 0.054) (Table 3).

3.4. Selection and Genetic Gain

3.4.1. Selection by Principal Component Method

The results of the principal component analysis (PCA) for growth and stem quality traits are shown in

Table 4. Eigenvalues and variance contributions of different components for progeny families.

Component	Eigenvalue	Variance Contribution	Cumulative Variance Contribution	Normalized Weight Value
I	4.303	0.538	0.538	0.633
II	1.810	0.226	0.764	0.266
III	0.688	0.086	0.850	0.101
IV	0.457	0.057	0.907	−
V	0.357	0.045	0.952	−
VI	0.148	0.019	0.971	−
VII	0.143	0.018	0.989	−
VIII	0.092	0.012	1.000	−

and

Table 5. Factor and component loading values of different traits in the extracted components.

Trait	Component I		Component II		Component III
	Factor Loading Value	Component Loading Value	Factor Loading Value	Component Loading Value	Factor Loading Value	Component Loading Value
PIH	0.77037	0.46406	0.28323	−0.07874	0.22149	0.10135
PIBD	0.86869	0.47752	0.19681	−0.19135	0.10992	0.01304
CD	0.87772	0.45868	0.18826	−0.23655	−0.10431	−0.19287
SS	0.09517	0.16060	0.0876	0.17901	0.94400	0.88731
NW	−0.03813	0.21880	0.82213	0.58163	0.22131	−0.01466
NB	0.20434	0.32716	0.81520	0.47329	0.16272	−0.08026
BD	0.81759	0.34255	−0.18813	−0.4362	0.01075	0.02725
BIA	0.13634	0.21955	0.66728	0.34149	−0.20962	−0.39703

PIH, periodic increment in height; PIBD, periodic increment in basal diameter; CD, crown diameter; SS, stem straightness; NW, number of whorls; NB, number of branches; BD, branch diameter; BIA, branch insertion angle.

. Principal Components I, II, and III, with eigenvalues of 4.303, 1.820, and 0.688, respectively, accounted for a cumulative 85.0% of the total variance, indicating that the majority of the variation in the evaluated traits is explained by these three components (Table 4). In Component I, the highest factor loadings (0.77037 to 0.87772) corresponded to crown diameter, periodic increment in basal diameter, branch diameter, and periodic increment in height. The number of branches and number of whorls, with factor loadings of 0.81520 and 0.82213, respectively, mainly defined component II. Component III was primarily associated with stem straightness, which had the highest loading value (0.944) (Table 5).

Table 6. Standardized values and principal component values of growth traits and stem quality of 16 (25%) best-ranked families.

Family	Standardized Value								Component Value			PCV	Rank
	PIH	PIBD	CD	SS	NW	NB	BD	BIA	I	II	III
Chi20	1.754	0.959	0.798	1.788	0.545	1.176	−0.838	2.866	2.338	−0.451	0.611	1.421	1
Yol46	2.388	1.731	1.100	0.528	−0.073	0.525	0.259	−0.268	2.221	−0.869	0.662	1.241	2
Ixt01	1.231	1.361	2.462	1.013	0.545	0.123	1.634	0.638	1.934	−0.203	0.287	1.199	3
Teo40	1.499	1.402	1.190	0.480	0.100	−0.091	1.538	1.006	1.690	0.418	−0.081	1.172	4
Chi04	1.085	1.078	0.673	0.503	−0.275	0.046	0.984	0.054	1.128	0.225	0.392	0.813	5
Teo44	0.216	1.345	−0.179	0.273	−1.447	−1.169	1.544	−1.544	0.535	1.358	0.773	0.778	6
Jal20	0.998	−0.179	0.135	0.528	−0.197	−1.436	0.400	−1.113	0.656	0.606	0.855	0.663	7
Ixt09	0.871	0.380	0.806	0.962	−0.158	0.784	0.113	0.577	0.976	−0.193	0.620	0.629	8
Ixt08	0.208	1.612	1.048	0.262	−0.679	0.158	1.603	−0.528	0.821	0.313	0.241	0.627	9
Ixt12	0.231	0.949	0.437	0.044	−0.568	0.135	−0.658	0.155	1.107	−0.263	−0.051	0.625	10
Ver04	−0.152	0.003	−0.339	0.916	−1.013	0.113	0.325	1.422	0.308	1.419	0.283	0.601	11
Jal14	0.408	−0.114	0.881	−1.334	−1.095	−0.168	0.719	1.644	0.734	1.134	−2.015	0.562	12
Yol45	0.177	−0.299	0.307	−0.174	−1.013	−0.521	0.258	0.267	0.415	0.979	−0.370	0.485	13
Jal17	−0.165	0.227	0.458	−0.199	−0.815	−0.908	−0.960	−1.253	0.739	−0.117	0.161	0.453	14
Teo41	−0.418	0.281	−0.083	0.589	−0.568	−0.770	0.449	−0.083	0.201	0.966	0.451	0.430	15
Teo31	1.041	0.633	0.341	1.352	0.100	0.814	0.751	−0.364	0.534	−0.282	1.439	0.408	16

PIH, periodic increment in height; PIBD, periodic increment in basal diameter; CD, crown diameter; SS, stem straightness; NW, number of whorls; NB, number of branches; BD, branch diameter; BIA, branch insertion angle; PCV, principal component value.

shows the top 16 families, ranked in descending order based on their principal component values; these families represent the top 25% of the 64 evaluated families. The complete ranking of all families (from 1 to 64), along with the equations used to calculate the principal components for growth and stem quality traits, is provided in Table S4.

3.4.2. Selection by Multi-Trait Comprehensive Evaluation Method

Table 7. Trait evaluation value and comprehensive evaluation value of 16 families (25%) best-ranked by the multi-trait comprehensive evaluation method.

Family	Trait Evaluation Value								CEV	Rank
	PIH	PIBD	CD	SS	NW	NB	BD	BIA
Chi20	0.939	0.933	0.846	0.944	−0.862	−0.833	−0.797	1.000	1.473	1
Yol46	1.000	1.000	0.874	0.864	−0.814	−0.777	−0.884	0.779	1.429	2
Ixt01	0.888	0.968	1.000	0.895	−0.862	−0.743	−0.993	0.843	1.413	3
Teo40	0.914	0.971	0.883	0.861	−0.828	−0.725	−0.985	0.869	1.400	4
Ixt12	0.791	0.932	0.813	0.833	−0.776	−0.744	−0.812	0.809	1.359	5
Chi04	0.874	0.943	0.835	0.863	−0.799	−0.737	−0.941	0.802	1.356	6
Ixt09	0.853	0.882	0.847	0.892	−0.808	−0.800	−0.873	0.839	1.354	7
Jal14	0.808	0.839	0.854	0.746	−0.735	−0.718	−0.920	0.914	1.337	8
Chi10	0.972	0.901	0.739	1.000	−1.000	−1.000	−0.731	0.897	1.334	9
Jal17	0.752	0.868	0.815	0.818	−0.757	−0.655	−0.788	0.709	1.328	10
Ver04	0.754	0.849	0.741	0.889	−0.741	−0.742	−0.889	0.898	1.326	11
Jal20	0.865	0.833	0.785	0.864	−0.805	−0.610	−0.895	0.719	1.326	12
Ixt08	0.789	0.990	0.869	0.847	−0.767	−0.746	−0.990	0.761	1.324	13
Ixt03	0.897	0.955	0.848	0.895	−0.872	−0.808	−0.889	0.713	1.319	14
Yol45	0.786	0.822	0.801	0.819	−0.741	−0.688	−0.884	0.817	1.316	15
Jal13	0.698	0.827	0.810	0.778	−0.733	−0.661	−0.815	0.824	1.314	16

PIH, periodic increment in height; PIBD, periodic increment in basal diameter; CD, crown diameter; SS, stem straightness; NW, number of whorls; NB, number of branches; BD, branch diameter; BIA, branch insertion angle; CEV, Comprehensive evaluation value.

lists the top 16 families, ranked in descending order based on their comprehensive evaluation scores. The results of the multi-trait comprehensive evaluation for all families are provided in Table S5.

3.4.3. Genetic Gain with Three Selection Intensities

With a selection intensity of 6.25% (four families selected out of sixty-four), both evaluation methods identified the same families, and the genetic gains were identical. Genetic gains for growth traits ranged from 8.01% to 11.92%. For stem quality traits, gains were observed only for stem straightness (2.12%) and branch insertion angle (4.78%), while no genetic gains were obtained for the remaining traits (

Table 8. Genetic gain [%] and realistic gain [%] for growth and stem quality traits obtained with three selection intensities and two evaluation methods.

Traits	Principal Component Method			Multi-Trait Comprehensive Evaluation Method
	Genetic Gain	Realistic Gain	Family Selected	Genetic Gain	Realistic Gain	Family Selected
	Selection intensity 6.25% (4 families)
Periodic Increment In Height	11.92	21.68	Chi20, Yol46, Ixt01, Teo40	11.92	21.68	Chi20, Yol46, Ixt01, Teo40
Periodic Increment In Basal Diameter	8.01	14.05		8.01	14.05
Crown Diameter	10.44	16.57		10.44	16.57
Stem Straightness	2.12	7.30		2.12	7.30
Number of Whorls	0.00	0.00		0.00	0.00
Number of Branches	0.00	0.00		0.00	0.00
Branch Diameter	0.00	0.00		0.00	0.00
Branch Insertion Angle	4.78	9.37		4.78	9.37
	Selection intensity 12.5% (8 families)
Periodic Increment In Height	8.71	15.84	Chi20, Yol46, Ixt01, Teo40, Chi04, Teo44, Jal20, Ixt09	8.21	14.93	Chi20, Yol46, Ixt01, Teo40, Ixt12, Chi04, Ixt09, Jal14
Periodic Increment In Basal Diameter	5.93	10.41		5.69	9.98
Crown Diameter	6.57	10.43		7.85	12.46
Stem Straightness	1.69	5.82		1.11	3.82
Number of Whorls	0.58	1.13		0.59	1.15
Number of Branches	0.03	0.06		0.00	0.00
Branch Diameter	0.00	0.00		0.00	0.00
Branch Insertion Angle	1.25	2.45		3.76	7.37
	Selection intensity 25% (16 families)
Periodic Increment In Height	4.45	8.09	Chi20, Yol46, Ixt01, Teo40, Chi04, Teo44, Jal20, Ixt09, Ixt08, Ixt12, Ver04, Jal14, Yol45, Jal17, Teo41, Teo31	5.74	10.43	Chi20, Yol46, Ixt01, Teo40, Chi04, Jal20, Ixt09, Ixt08, Ixt12, Ver04, Jal14, Yol45, Jal17, Chi10, Ixt03, Jal13
Periodic Increment In Basal Diameter	4.08	7.15		3.92	6.88
Crown Diameter	4.63	7.36		5.08	8.07
Stem Straightness	0.85	2.94		1.13	3.91
Number of Whorls	2.30	4.51		0.85	1.67
Number of Branches	1.01	1.84		0.00	0.00
Branch Diameter	0.00	0.00		0.00	0.00
Branch Insertion Angle	0.18	0.36		1.70	3.33

).

With a selection intensity of 12.5% (eight families selected), six families were selected by both methods, while two differed between the methods (Table 8). The genetic gains estimated by both methods were similar, though lower than those obtained with a 6.25% selection intensity. For growth traits, genetic gains ranged from 5.69% to 8.71%. For stem quality traits, gains ranged from 0.0 to 3.76%. At a selection intensity of 25% (16 families selected), 13 families were selected by both methods, and 3 differed in each method. Genetic gains for growth traits ranged from 3.92% to 5.74%, while gains for stem quality traits ranged from 0.18% to 2.30%, except for branch diameter, which showed no gain under either method (Table 8). Realistic gains were 59% to 96% higher than the genetic gains, except for stem straightness, where the realistic gain was 245% higher (Table 8).

4. Discussion

4.1. Differences Between Families

The growth in basal diameter, height, and crown diameter observed in this study was similar to that reported for two varieties of Gregg pine (Pinus greggii Engelm. ex Parl.) planted at two sites within the same region [35]. In contrast, progenies of austral Gregg pine (Pinus greggii var. australis Donahue & López Upton) showed lower diameter growth and comparable height growth [14]. Although no studies have been conducted on smooth-bark Mexican (Pinus pseudostrobus Lindl. var. pseudostrobus) varieties in this region, research in Michoacán, Mexico, has reported lower diameter and height growth for both Apulco pine (Pinus pseudostrobus var. apulcensis) [36] and smooth-bark Mexican [37] compared to the present study. These results highlight the potential of Apulco pine for establishing productive plantations and contributing to soil recovery in the region—a reasonable expectation given that this variety is native to the region [17]. From a soil protection perspective, crown diameter, number of whorls, and number of branches are among the most important traits [38]; higher values for these characteristics contribute to increased biomass production, which can be incorporated into the soil [38,39,40]. However, these variables are positively correlated with stem diameter, height, and volume growth, as trees with greater foliar area exhibit higher photosynthetic capacity [38]. In line with this, the observed high growth rates may be attributed to large crown dimensions and adequate site fertility at the experimental location [19].

The wide phenotypic variation observed among families in stem quality traits is relevant because these traits directly influence wood quality [41,42]. Phenotypes with more twisted stems, a higher number of whorls and branches, thicker branches, and lower branch insertion angles tend to reduce wood quality [43,44]. Currently, no information is available for these traits in Apulco pine and smooth-bark Mexican. However, the number of whorls observed in this study was lower than that reported for Gregg pine provenances tested in the Mixteca Oaxaqueña region [35]. Stem straightness, number of whorls, and branch insertion angle were similar to those of austral Gregg pine progenies planted at two sites in the same region [14]. In contrast, branch diameter was greater than that of austral Gregg pine [14].

The percentage contribution of family to the total variance suggests that early family selection is feasible to increase genetic gain in growth and stem quality traits. The family contribution to total variance observed in this study was higher than that reported for growth traits (1%–9%) in progeny trials of egg-cone pine (Pinus oocarpa Schiede ex Schltdl.) and red pine (Pinus patula Schl. et Cham.) of the same age [13,45]. In contrast, for stem quality traits, the family contribution was similar to previously reported values (1%–15%) for same-aged red pine and egg-cone pine plants [13,45].

4.2. Genetic Control

The genetic variation values obtained in this study suggest significant differences among families for most of the traits evaluated [46], indicating that genetic gains can be achieved through selection [47]. The coefficients of genetic variation for growth traits observed in this study were higher than those reported for red pine, austral Gregg pine and loblolly pine (Pinus taeda L.) (7.76 to 16.11%) [14,26,48]. Likewise, lower values have been reported for red pine (CVg = 6.19%–6.52%), egg-cone pine (2.7%–6.3%), and Douglas fir (Pseudotsuga menziesii var. menziesii (Mirb.) Franco) (3.21%–3.60%) [13,49,50].

The coefficients of genetic variation for stem quality traits in Apulco pine were similar to those reported for the same traits in red pine and austral Gregg pine (CVg = 7.06%–15.93%) [14,26], except for branch insertion angle in austral Gregg pine (CVg = 3.86%–6.69%) and number of branches in red pine (CVg = 2.97%) [14,26]. Lower genetic variation has also been reported for branch number, diameter, and insertion angle (CVg = 3.09%–4.09%) in Douglas fir [50], and for the number of whorls in red pine (CVg = 4.27%–7.80%) [49]. The coefficient of genetic variation for stem straightness in Apulco pine was higher than that of white pine (Pinus strobus L.) × Himalayan blue pine (Pinus wallichiana A.B.Jacks.) hybrids (CVg = 2.1%) [51] but lower than the value reported for radiata pine (Pinus radiata D.Don) (CVg = 25.2%) [52]. Differences in genetic variation between the present and previous studies may be attributed to variation in age at the time of evaluation. For instance, in loblolly pine, the coefficient of genetic variation for growth traits decreased with age, from 14% at 1.5 years to 2.94% at 22.5 years, while for stem straightness, it increased from 1.51% to 5.74% [48].

The individual heritability ($h_i^2$) and family mean heritability ($h_f^2$) for growth traits were classified as high [14,26,47], indicating strong genetic control over these traits. The high heritability values found in this study were expected, as Apulco pine has also shown high values for early growth in diameter ($h_i^2$ = 0.26–0.55; $h_f^2$ = 0.87–0.93) and height ($h_i^2$ = 0.40–0.95; $h_f^2$ = 0.60–0.98) [20,53]. The $h_i^2$ and $h_f^2$ values observed here were higher than those reported for diameter and height in austral Gregg pine in the Mixteca Oaxaqueña region ($h_i^2$ = 0.07–0.11; $h_f^2$ = 0.21–0.30) [14]. Other Pinus species have shown a wide range of heritability estimates for diameter ($h_i^2$ = 0.00–0.30; $h_f^2$ = 0.07–0.72) and height ($h_i^2$ = 0.0014–0.73; $h_f^2$ = 0.12–0.84) [13,14,21,26,48,49,54,55,56,57,58,59], within which the heritability values from this study are also included.

The heritability estimates ($h_i^2$ and $h_f^2$) for stem quality traits obtained in this study were classified as moderate to high [14,26,47], except for stem straightness, which showed low $h_i^2$ and moderate $h_f^2$. These results indicate that most stem quality traits in Apulco pine are under strong genetic control, whereas stem straightness appears to be influenced more by environmental factors than by genetic ones.

Information on the genetic control of traits influencing stem quality is lacking for both varieties of on smooth-bark Mexican. In the Mixteca Oaxaqueña region, where this study was conducted, the heritability of stem straightness in austral Gregg pine was slightly higher ($h_i^2$ = 0.11 and 0.14; $h_f^2$ = 0.31 and 0.34) than the values obtained for the same trait in the present study. In contrast, the heritability values for the number of whorls ($h_i^2$ = 0.15 and 0.18; $h_f^2$ = 0.37 and 0.39), branch diameter ($h_i^2$ = 0.06 and 0.09; $h_f^2$ = 0.21 and 0.26), and branch insertion angle ($h_i^2$ = 0.14 and 0.07; $h_f^2$ = 0.35 and 0.21) in austral Gregg pine were lower than those found for Apulco pine [14].

In maritime pine (Pinus pinaster Aiton), individual heritability ($h_i^2$ = 0.05–0.15) and family mean heritability ($h_f^2$ = 0.32–0.61) for stem straightness were similar and higher, respectively, than those obtained in this study. However, in loblolly pine, a wider range of individual heritability values was reported ($h_i^2$ = 0.01–0.83), which tended to increase with the age of the trials [48,54].

In general, other Pinus species have shown lower heritability estimates for number of whorls ($h_i^2$ = 0.09 to 0.18; $h_f^2$ = 0.206 to 0.36), number of branches ($h_i^2$ = 0.00 to 0.17; $h_f^2$ = 0.00 to 0.47), branch diameter ($h_i^2$ = 0.04 to 0.08; $h_f^2$ = 0.10 to 0.42), and branch insertion angle ($h_i^2$ = 0.07 to 0.35; $h_f^2$ = 0.26 to 0.47) [21,26,48,49,54]], with the exception of maritime pine, which showed higher genetic control for branch insertion angle ($h_f^2$ = 0.52 to 0.79) [55].

High genetic control over growth traits in Apulco pine has positive implications for genetic improvement, as it enables greater genetic gains and makes early selection more effective [7,60]. In contrast, the low genetic control of stem straightness suggests that genetic gains for this trait will be limited, early selection will be less effective, and more generations will be required for improvement. High heritability values observed in this study may be attributed to low mortality rates and pronounced differentiation among families [7,49], which is consistent with the levels of variability identified. Moreover, the relatively high heritability values compared to those reported for other Pinus species suggest that environmental influence on traits of Apulco pine is minimal [26], possibly due to the low environmental heterogeneity at the planting site. Environmental variability can alter heritability estimates through genotype-by-environment interactions [14,54]. Differences from other Pinus species may also result from evaluation age and growing conditions [49]. In several Pinus species, genetic control of growth and wood quality traits tends to increase with tree age [13,21,48,49,54,55,57]; therefore, heritability values in Apulco pine are also expected to increase over time.

4.3. Correlations Between Traits

Positive and high genetic correlations among growth traits of Apulco pine are common due to the inherent interrelationships in tree growth rates [26]. These strong correlations also suggest strong genetic association, possibly due to pleiotropy or linkage among closely located genes [14,25]. In the Mixteca Oaxaqueña region, where the present study was conducted, genetic correlations between growth traits of austral Gregg pine were also positive and high (r_g ≥ 0.77) [14]. Similarly, in other Pinus species, high genetic correlations have been reported among growth traits (r₉ = 0.47−0.97) [13,21,26,49,55,56,57].

The moderate to high positive genetic associations between periodic height and basal diameter increments with stem straightness, and between periodic height increment and branch insertion angle, represent clear advantages for genetic improvement, since it is possible to increase productivity while simultaneously achieving favorable responses in stem straightness and branch insertion angle. On the contrary, the genetic correlations between growth and the remaining stem quality traits were unfavorable for genetic improvement, as they suggest that trees that are more productive will have a greater number of branches and whorls, and thicker branches. For sawn wood, this would result in lower quality due to the higher number of larger knots and their lower insertion angle [61]. In the Mixteca Oaxaqueña region, the genetic correlations between growth traits and stem quality traits of Gregg pine austral Gregg pine were positive and weak (r_g = 0.02 to 0.12) with the number of whorls, weak to moderate (r_g = 0.03 to 0.40) with branch insertion angle, and weak to high (r_g = 0.09 to 0.97) with branch diameter; whereas correlations with stem straightness were negative and weak (r_g = −0.02 to −0.34) [14].

The genetic correlations between growth traits and stem straightness in austral Gregg pine [14] contrast with those observed for Apulco pine in this study. In general, genetic correlations between growth and stem quality traits in other Pinus species range from negative to positive and from weak to high (r_g = −0.59 to 0.84) [21,26,49,55]. Furthermore, this genetic association can vary within the same species. For example, in red pine, a weak to moderate negative genetic correlation (r_g = −0.28 to −0.49) between growth traits and the number of whorls was reported in some trials [21,26]; whereas, in another study, a weak to strong positive association was found (r_g = 0.23 to 0.89) [49]. Site conditions may also influence genetic associations, as observed in austral Gregg pine, where the genetic correlation between stem diameter and stem straightness was positive at one site but negative at another [14].

All genetic correlations among stem quality traits were unfavorable from a breeding perspective, since selection for improvement based on one trait tends to result in reduced stem quality when considering the others. The only favorable genetic correlations were those between stem straightness and branch diameter (negative) and between stem straightness and branch insertion angle (positive). This indicates that selecting for straighter trees will also result in thinner branches and wider insertion angles, which is associated with higher wood quality [62]. In the Mixteca Oaxaqueña region, genetic correlations among stem quality traits of austral Gregg pine ranged from negative to positive (r_g = −0.67 to 0.69) [14], consistent with the results reported for other Pinus species [21,26,55]. However, the genetic correlations between the same pairs of traits varied across sites. For example, the correlation between branch insertion angle and the number of whorls, as well as with branch diameter, shifted from positive to negative depending on the site [14]. This suggests an environmental influence on genetic correlation responses, likely due to genotype–environment interactions [35,36,46]. These findings highlight the importance of evaluating Apulco pine families at additional sites within the Mixteca Oaxaqueña region.

In general, phenotypic correlations followed the same pattern as genetic correlations, except for crown diameter, which showed positive phenotypic correlations with all traits but negative genetic correlations with several. However, in cases where the directions of the correlations did not match, the phenotypic correlation was not statistically significant. This may be attributed to the age at the time of evaluation. For example, in loblolly pine, the direction of phenotypic and genetic correlations between stem straightness and growth traits did not align at 1.5 years of age but did align at 9.5, 13.5, and 22.5 years [48]. Discrepancies between the direction of phenotypic and genetic correlations have also been reported in austral Gregg pine (e.g., stem conformation and straightness vs. height, diameter, volume, and branch diameter) [14], red pine (e.g., branch diameter vs. height, wood density, and volume; branch insertion angle vs. wood density) [21,26], and Scots pine (Pinus sylvestris L.) (e.g., fiber width vs. stem diameter and wood density; fiber length vs. diameter; height vs. wood density) [56].

On the other hand, it is important to highlight that, unlike phenotypic correlations, the majority (82%) of genetic correlations were not statistically significant, which was consistent with the high standard error values. The stressful environment caused by low soil moisture at the experimental site may have increased environmental variance and limited the phenotypic expression of the families’ genetic potential, thereby reducing genetic correlations [7,63]. Additionally, the early age at which the evaluation was conducted may have limited the accurate estimation of genetic correlations. At the juvenile stage, traits such as stem straightness, number of whorls, branch diameter, and branch insertion angle may not have been fully expressed, and it is likely that environmental influences predominated over genetic expression—possibly explaining the low heritability observed for these traits [25]. Moreover, environmental heterogeneity within the experimental site, combined with potential residual error, may have reduced the precision of genetic covariance estimates, thereby hindering the detection of statistically significant genetic correlations [7].

4.4. Selection and Genetic Gain

Evaluation methods for selecting forest trees follow different principles and rationales; however, they all aim to increase genetic gain, whether for individual traits or for multiple traits of economic or adaptive interest [7,64]. Principal component and multi-trait comprehensive evaluation methods are used to select families and clones based on multiple traits [30,31]. The effectiveness of these methods was demonstrated in this study, as both methods selected the same families and achieved the same genetic and realistic gains for all traits at a selection intensity of 6.25%. Similarly, at selection intensities of 12.5% and 25%, both methods coincided in selecting most of the same families, differing only in two and three families, respectively. The efficiency and consistency of these methods in identifying the same families have also been reported in other studies [30,31].

At selection intensities of 12.5% and 25%, slight differences were observed between the two evaluation methods in terms of the range of selected families and in the genetic and realistic gains obtained—findings that are consistent with previous studies [30,31]. These differences arise because the comprehensive multi-trait evaluation method applies equal standardized weights to each trait, whereas the principal component method transforms correlated traits into a set of uncorrelated variables through orthogonal transformation and subsequently evaluates each family based on its principal component score [30,31].

The four families selected (Chi20, Yol46, Ixt01, and Teo40) by both evaluation methods at a 6.25% selection intensity should be considered elite under the environmental conditions of the test site in the Mixteca Oaxaqueña region. The extensive use of these families in commercial plantation programs in the region could ensure higher productivity and improved stem quality. However, due to the limited number of families selected, genetic diversity would be reduced [7,12]. In breeding programs, maintaining high genetic diversity is essential to ensure adaptability and resilience to changing environmental conditions [6,7,12].

Therefore, although genetic gain would be lower, it is recommended to broaden the genetic base in timber plantation programs by using eight or sixteen families (corresponding to selection intensities of 12.5% and 25%, respectively), as listed in Table 8. In contrast, for reforestation programs, a larger number of families should be used to ensure the conservation of genetic diversity [65]. The family rankings obtained in this study provide a valuable reference for decision-makers when determining the appropriate number of families to include in reforestation efforts.

The number of families propagated in operational plantations is a critical factor: a smaller number results in higher genetic gain but reduced genetic diversity, whereas a larger number leads to lower genetic gain but enhanced genetic diversity [7]. The balance between genetic gain and diversity depends on multiple factors; however, variation is often influenced by geographic isolation [30]. Therefore, selecting families from different provenances and broader geographic distances can help to increase genetic diversity [12].

Based on the above, the 16 families selected in this study (at a 25% selection intensity) are likely to exhibit broad genetic diversity, as they include families from all eight provenances and the three geographic regions. Moreover, families from early generation selection populations (with low levels of genetic improvement) tend to retain a broad genetic base and structure [12]. Ideally, family selection should be complemented by molecular genetic diversity studies [12,66] to assess how diversity changes under different selection intensities. In Mexico, such studies are scarce; however, in egg-cone pine, it was shown that selecting between 2 and 15 individuals from natural stands did not reduce genetic diversity [12]. Another approach not explored in this study but worth considering in future genotype selection for the Mixteca Oaxaqueña region is individual or combined selection, as these methods often offer advantages over family selection in terms of both genetic gain and genetic diversity [12,67,68].

In this study, with a selection intensity of 12.5%, the genetic gain in height was higher and lower in diameter than the gains reported for Korean pine (Pinus koraiensis Siebold & Zucc.) at a selection intensity of 10% (4.85% for height and 9.96% for diameter) [31]. In contrast, the realistic gains obtained by both methods at a 12.5% selection intensity were lower than those reported for Yunnan pine (Pinus yunnanensis French.) in height (23.42%–26.89%), diameter (16.75%–18.29%), and crown diameter (14.99%–17.19%) with a 10% selection intensity [30]. Although the genetic gains in growth traits were moderate, the realistic gains were greater, which is highly beneficial for genetic improvement and for producing higher-quality germplasm [30] for reforestation and plantation programs in the Mixteca Oaxaqueña region.

In growth traits, stem straightness, and branch insertion angle, the genetic and realistic gains estimated by both selection methods decreased as the selection intensity increased from 6.25% to 12.5% and 25%. In contrast, for the number of whorls and the number of branches, both genetic and realistic gains increased. This pattern reaffirms the genetic and unfavorable phenotypic relationships that the latter traits have with the former. Moreover, it highlights the difficulty of simultaneously improving growth and stem quality traits (i.e., fewer whorls, fewer branches, and smaller branch diameter).

Among these traits, no genetic gain was obtained for branch diameter at any selection intensity, underscoring the greater challenge of improving this trait in conjunction with growth. A similar pattern has been observed in other pine species; for example, in Yunnan pine, with a selection intensity of 10%, a high realistic gain was achieved for height (26.54%), but only a marginal gain (0.02%) for the number of branches [30].

Finally, the importance of early selection of Apulco pine families is emphasized, given the urgent need for high-productivity families with good wood quality in the Mixteca Oaxaqueña region. Following the successful reforestation of degraded areas [16,17,18], landowners now seek more productive plantations that yield high-quality timber. Although early selection is sometimes questioned due to the long rotation periods typical of the Pinus genus, studies on smooth-bark Mexican, Apulco pine and other pine species have shown high age–age genetic correlations for diameter, height, and stem quality traits [20,21,22], supporting the feasibility of early selection. Therefore, the information generated in this study should serve as a basis for stakeholders interested in establishing commercial forest plantations in the region. Future evaluations will be necessary to confirm the findings of this study, explore alternative selection methods, and complement them with molecular genetic diversity analyses.

5. Conclusions

The growth and stem quality of open-pollinated families of Apulco pine (Pinus pseudostrobus var. apulcensis) varied due to genotypic differences. The genetic control of growth traits was high, indicating strong potential for substantial genetic gains through selection. Genetic correlations among growth traits were favorable for breeding purposes; however, genetic correlations among stem quality traits, and between growth and stem quality traits, were not significant, posing challenges for the simultaneous improvement of both trait types.

Both evaluation methods proved effective and consistent in selecting superior families of Apulco pine based on growth and stem quality traits. To increase productivity in commercial forest plantations in the Mixteca Oaxaqueña region, it is recommended that forest managers use the 16 selected families. For reforestation purposes, a larger number of families should be used to ensure greater genetic diversity.