Genetic Variation and the Relationships Among Growth, Morphological, and Physiological Traits in Pterocarpus macrocarpus: Implications for Early Selection and Conservation

Abstract

Understanding genetic variation in commercially valuable tree species is essential for improving breeding and conservation efforts. This study investigates genetic variation, heritability, and trait relationships in Pterocarpus macrocarpus, a vital hardwood species for Thailand’s reforestation initiatives. We evaluated growth (height and diameter), morphology (biomass dry weight and specific leaf weight), and physiological traits (net photosynthesis [A], transpiration rate [E], and water-use efficiency [WUE]) across 112 open-pollinated families from six natural populations under controlled nursery conditions over 30 weeks. Using a randomised complete block design, variance and covariance analyses were conducted to estimate genetic parameters. Seedling survival reached 95%, confirming favourable conditions for genetic expression. There were significant differences among populations and families within populations in growth and biomass. In contrast, physiological traits showed notable family-level variation (A, E, WUE) and only population effects for WUE. Residual variance was predominant across traits, indicating considerable within-family variation. Growth and biomass exhibited moderate to high heritability (individual: 0.39–1.00; family: 0.61–0.90), while specific leaf weight and shoot-to-root ratio had lower heritability at the individual level. Physiological traits showed low to moderate heritabilities (individual: 0.26–0.43; family: 0.47–0.62), with maternal effects via seed weight significantly influencing early growth. The heritability of height decreased over time, whereas the heritability of diameter remained stable. Strong genetic correlations among growth and biomass suggest the potential for combined selection gains. However, physiological traits show weak or no correlations with growth, highlighting their independent genetic control. Variation at the population level in growth and WUE may reflect adaptive responses to seed-source environments. Our findings support the use of nursery-based screening as a cost-effective method for the early identification of high-quality families. WUE is a promising focus for breeding programs targeting drought-prone regions. This study provides key insights for advancing the genetic improvement and conservation of P. macrocarpus, emphasizing the importance of incorporating physiological traits into breeding and conservation strategies.

1. Introduction

Genetic diversity is crucial for forest resilience, enabling tree species to adapt to environmental changes, combat pests and diseases, and maintain ecosystem functions amid global shifts [1,2]. In an era of rapid deforestation, habitat fragmentation, and climate instability, understanding patterns of genetic variation has become vital for protecting biodiversity and enhancing the productivity and sustainability of reforestation efforts [3]. Genetic variation occurs at multiple levels—distributed among populations due to different selection pressures, founder effects, or genetic drift, and within populations as a result of mutation, recombination, and balancing selection [4]. The balance of these factors guides breeding and conservation strategies: high divergence between populations suggests the need for provenance-based approaches to preserve locally adapted gene pools. Conversely, significant variation within populations emphasises the importance of within-population selection in identifying superior genotypes. However, genetic architecture is dynamic, changing across traits, developmental stages, and environmental contexts [5]. For example, growth may be highly heritable in saplings but becomes increasingly influenced by ecological variability as trees mature [6]. This complexity requires a comprehensive understanding of trait heritability and plasticity to develop breeding and conservation programs that can withstand climate uncertainties.

While significant progress has been made in quantifying the genetic control of growth and structural traits (e.g., height, stem form, wood density), physiological traits such as photosynthetic efficiency, stomatal regulation, and nutrient-use efficiency remain disproportionately understudied in forestry research [7,8]. This knowledge gap results from historical biases toward easily measurable field traits, logistical challenges in scaling physiological measurements, and misconceptions that physiological processes are too environmentally variable to serve as reliable selection criteria [9,10]. Nevertheless, physiological traits act as the mechanistic link between genotype and phenotype, directly influencing a tree’s ability to optimise carbon gain, reduce water stress, and adapt to abiotic extremes [11]. Recent advances in phenotyping technology and growing awareness of climate-driven selection pressures have renewed interest in physiological traits as early indicators of adaptive potential [12,13]. For example, water-use efficiency (WUE), which assesses the balance between carbon assimilation and transpiration, has become a key trait in breeding drought-resilient genotypes for water-limited ecosystems [14]. However, integrating physiological traits into operational breeding programs remains in the early stages, hindered by a lack of empirical data on their heritability and genetic relationships with productivity.

Pterocarpus macrocarpus Kurz (Fabaceae), a nitrogen-fixing hardwood endemic to the seasonal tropical forests of mainland Southeast Asia, exemplifies a species for which bridging this knowledge gap is both ecologically urgent and socioeconomically strategic. Renowned for its dense, termite-resistant timber and ornamental value, P. macrocarpus is a keystone species in Thailand’s reforestation efforts, where it is widely planted to restore degraded watersheds, combat soil erosion, and create livelihood opportunities through community forestry [15,16]. Its ecological versatility, which enables it to tolerate nutrient-poor soils, seasonal drought, and fire, makes it a candidate for climate-smart reforestation. However, despite its importance, the species lacks structured breeding programs, and current genetic studies are limited to assessments of population diversity using neutral markers. Isozyme analyses reveal significant genetic differentiation among Thai populations, with a distinct east–west trend likely influenced by historical biogeographic barriers and environmental gradients [17]. Yet, neutral markers provide limited insight into adaptive genetic variation or the heritability of traits vital to reforestation success. This gap highlights the need for trait-based studies to unlock the species’ full potential in restoration ecology.

Nursery trials offer a practical yet often overlooked platform for advancing research. By providing controlled growing conditions, nurseries reduce environmental variability, making genetic differences among families and populations more visible than in diverse field environments [18]. This setting is especially helpful for assessing physiological parameters like photosynthetic rates and WUE, which are sensitive to microenvironmental changes and are hard to measure on a large scale in natural stands [14,19,20]. Early-life trait expression in nurseries also allows for quick screening, avoiding the decades-long wait for maturity in slow-growing hardwoods. Although age–age correlations between nursery and field performance can weaken due to shifting selection pressures across ontogeny [21], nursery data remain essential for excluding poorly adapted families before incurring the costs of lengthy, long-term field trials [22]. Furthermore, nursery-based heritability estimates provide baseline parameters for predicting genetic gains and improving breeding strategies, especially for species like P. macrocarpus, where no established breeding infrastructure exists.

In this study, we address these scientific and practical gaps by performing a comprehensive genetic assessment of P. macrocarpus under nursery conditions. Using open-pollinated progeny from 112 families across six natural populations in Thailand, we (1) evaluate genetic variation in height, diameter, biomass allocation, specific leaf weight, net photosynthesis, transpiration, and water-use efficiency; (2) estimate heritabilities and maternal effects; and (3) analyse genetic correlations among traits to identify potential selection trade-offs or synergies. Incorporating physiology into a traditional growth-morphology framework improves a holistic understanding of P. macrocarpus’s genetic structure, providing practical insights for breeding programs focused on enhancing productivity and climate resilience in Southeast Asia’s vulnerable tropical forests.

2. Materials and Methods

2.1. Plant Material and Experimental Design

2.1.1. Seed Sources and Genetic Background

Open-pollinated seeds of P. macrocarpus were collected from 112 maternal families across six natural populations in Thailand (Figure 1). These populations were chosen to represent geographic and climatic diversity, covering different forest types (deciduous dipterocarp, mixed deciduous, and dry evergreen) and various environmental gradients (

Table 1. Geographic location, elevation, climatic information, forest type, inbreeding coefficient (F), multilocus outcrossing rate (t_m), and number of families of six populations of P. macrocarpus included in the nursery trial.

Population		Latitude (°N)	Longitude (°E)	Elevation (m)	Mean Annual Temperature (°C)	Annual Rainfall (mm)	Forest Type a	F b	tm b	No. of Families
No.	Name
1	Lampang	18°35′	99°54	350	25.9 c	1076 c	MDF	−0.010	0.945	15
2	Phuphan-3	16°58′	103°45	310	26.1 c	1587 c	DDF	0.144	0.751	14
3	Khong-chiam	15°24′	105°29′	200	26.7 c	1634 c	DDF	0.219	0.895	13
4	Uthaithani	15°30′	99°22′	220	N/A	N/A	DDF	0.090	0.953	21
5	Saraburi	14°35′	101°12′	200	26.1 d	1168 d	MDF	0.108	0.898	22
6	Sakaerat	14°25′	101°45′	380	26.3 e	1310 e	DDF	0.048	0.947	27

^a MDF: mixed deciduous forest (moist site); DDF: dry dipterocarp forest (dry site). ^b Outcrossing estimates [23]. ^c Meteorological Department, Bangkok, Thailand. ^d Thai-Danish Dairy Farm, Saraburi, Thailand. ^e Sakaerat Environmental Research Station, Nakhonratchasima, Thailand. N/A: data not available.

). Inbreeding coefficients (F) and multilocus outcrossing rates (t_m) were taken from earlier parentage analyses [23] to describe the mating system (Table 1).

The climate data in Thailand is mainly gathered from meteorological stations in urban areas rather than in forest regions. As a result, rainfall data are not used to classify forest types. In Thailand, forest types are typically classified based on topography and species composition. MDF includes both evergreen and deciduous tree species, while DDF primarily consists of deciduous trees. In MDF, only deciduous trees shed their leaves during the dry season, whereas evergreen trees retain their foliage year-round. In DDF, nearly all trees lose their leaves during the dry season. Therefore, MDF has a greater ability to retain moisture than DDF. Whether a site is moist or dry ultimately depends on the type of forest.

2.1.2. Seed Preparation and Germination Protocol

Eighty seeds from each family were individually weighed to evaluate seed mass variability. To break physical dormancy, seeds were scarified with medium-grit sandpaper (P80 grade) to abrade the seed coat. Seeds were sown in 24-cell multi-purpose trays (cell dimensions: 5.1 cm diameter × 10.2 cm depth; 190 mL volume) filled with a sterile substrate of coconut husk fibre, coarse sand, and compost (2:1:1 v/v), adjusted to pH 7.5 with dolomitic limestone. Trays were arranged in a randomised family-block layout to reduce positional bias. Sowing took place in late May, coinciding with Thailand’s rainy season, and germination was observed daily. Germination (radicle emergence ≥ 2 mm) was completed within 10 days, and final germination rates were determined at 14 days after sowing.

2.1.3. Nursery Management and Transplanting

Seedlings were kept in a shaded nursery under controlled conditions of 27 °C (±2 °C), 80% relative humidity (RH), and 680 µmol m⁻² s⁻¹ of photosynthetically active radiation (PAR). At three weeks after sowing, each seedling received 0.5 g of 2-month controlled-release formula fertiliser (Osmocote 13-13-13) to promote early growth. After 12 weeks, 16 healthy seedlings per family (totalling 1792) were randomly selected and transplanted into 3 L terracotta pots (17.8 cm diameter × 20.3 cm depth) filled with the same substrate mixture, arranged in a randomised complete block design (RCBD) with 16 replicates (single-tree plots). Pots were spaced 20 cm apart to avoid canopy overlap and ensure even light exposure.

2.1.4. Fertilisation and Environmental Controls

Irrigation was applied every 48 h through drip lines to maintain substrate moisture at 60–70% of field capacity. An additional 10 g of 2-month controlled-release formula fertiliser (Osmocote 13-13-13) was supplied to each seedling at weeks 12 and 21 during the 30-week growth period. Environmental conditions were carefully monitored with HOBO data loggers (Onset Computer Corporation, Bourne, MA, USA), recording average daily temperatures of 27 °C (range: 24–30 °C), relative humidity (RH) of 80% (±5%), and photosynthetically active radiation (PAR) levels of 680 µmol m⁻² s⁻¹ (±50 µmol) during the 30-week experiment.

2.2. Trait Assessments

2.2.1. Growth and Morphology

Height and Diameter: Seedling height (cm) was measured weekly from the soil surface to the apical meristem using a rigid ruler (±1 mm precision), starting at week 3 post-sowing. Root collar diameter (mm) was measured at the soil interface with digital calipers (Mitutoyo Corp.; ±0.01 mm precision) beginning at week 12. Data were collected at 3-week intervals, resulting in 10 height and 7 diameter measurements per seedling.

Biomass Partitioning: At harvest (30 weeks), seedlings were carefully uprooted, washed to remove soil, and separated into leaves, stems, taproots (>2 mm diameter), and fibrous roots (<2 mm diameter). Tissues were oven-dried at 80 °C for 48 h until a constant mass was reached, and dry weights (g) were recorded using an analytical balance (Mettler Toledo; accuracy of ±0.001 g). Specific leaf weight (SLWT; g cm⁻²) was calculated by dividing leaf dry mass by total leaf area. Leaf area was measured with a portable laser area meter (CI-202, CID Bio-Science) on four fully expanded leaves per seedling, selected from the mid-canopy to minimise positional variability. Data recorded included leaf dry weight (LEAF), stem dry weight (STEM), taproot dry weight (TROOT), fibrous root dry weight (FROOT), total dry weight (TOTAL, i.e., LEAF + STEM + TROOT + FROOT), shoot portion dry weight (SHOOT, i.e., LEAF + STEM), and root portion dry weight (ROOT, i.e., TROOT + FROOT).

2.2.2. Physiological Traits

Gas exchange parameters were measured at week 27 using an infrared gas analyser (IRGA; CI-301PS, CID Bio-Science) with a 1 L leaf chamber. Two mature, sun-acclimated leaves per seedling were selected from the upper third of the canopy to represent photosynthetic capacity under light-saturated conditions. Measurements were conducted between 08:30 and 15:30 under ambient sunlight (PAR: 1026–2090 µmol m⁻² s⁻¹) and temperatures ranging from 25 to 34 °C. The net photosynthetic rate (A; µmol CO₂ m⁻² s⁻¹) and transpiration rate (E; mmol H₂O m⁻² s⁻¹) were recorded over 30 s intervals following a 90 s chamber equilibration. Instantaneous water-use efficiency (WUE; µmol CO₂ mmol⁻¹ H₂O) was calculated as A/E [24]. The measurement order was randomised across replications to minimise diurnal variability, and chamber conditions were calibrated hourly using a certified CO₂ reference gas (400 ppm).

2.3. Statistical and Genetic Analysis

Data were validated for normality using the Shapiro–Wilk test and checked for homogeneity of variance with Levene’s test before analysis using SAS 9.4 [25].

2.3.1. Growth and Morphology

An analysis of variance and covariance was conducted for height, diameter, and biomass using the following linear model:

Y_ijk = µ + R_i + P_j + RP_ij + F_k(P_j) + ε_ijk

(1)

where

• µ—grand mean;
• R_i—i^th replication effect, i = 1–16;
• P_j—j^th population effect, j = 1–6;
• RP_ij—effect of replication-by-population interaction;
• F_k(P_j)—k^th family effect within the j^th population, k = 1, 2, …, n (n ranges from 13 to 27);
• ε_ijk—residual error.

All model effects (e.g., family, population, block) were treated as random to reflect the hierarchical variance structures in the experimental design. Variance and covariance components were estimated using the Type I method (sequential sum of squares) with the Varcomp procedure in SAS 9.4 [25]. This approach divides the observed variation into parts attributable to random effects while calculating expected mean squares (EMS) and expected mean cross products (EMCPs).

Hypothesis testing for variance components relies on Satterthwaite’s approximation [26], which adjusts the denominator degrees of freedom in mixed models to consider unbalanced designs and unequal replication. This method is robust for nested random effects and helps prevent the inflation of Type I error in small-sample contexts.

To validate model assumptions, residuals were checked for normality using the Shapiro–Wilk test (p > 0.95) and for homoscedasticity with Levene’s test (p > 0.05). All analyses were conducted at α = 0.05, with significance levels adjusted for multiple comparisons using the Benjamini–Hochberg procedure in SAS 9.4 [25] where relevant.

The observed inbreeding coefficients (F = 0.15–0.32; Table 1) indicate moderate inbreeding at the population level, requiring adjustments to how variance components are interpreted. According to quantitative genetic theory for partially inbred populations [27], family variance was assumed to be about one-third of the additive genetic variance, due to increased homozygosity and shared maternal pedigree effects. This adjustment aligns with expectations for open-pollinated progeny of tropical trees with mixed mating systems. The narrow-sense heritabilities for individuals and families were calculated as follows:

$$\mathrm{Individual} \mathrm{heritability}, h_i^2={\displaystyle \frac{3 x {\sigma }_{f\left(p\right)}^2}{{\sigma }_{e }^2+ {\sigma }_{f\left(p\right)}^2 }}$$

(2)

$$\mathrm{Family} \mathrm{heritability}, h_f^2={\displaystyle \frac{{\sigma }_{f\left(p\right)}^2}{\raisebox{1ex}{${\sigma }_e^2$}\!\left/ \!\raisebox{-1ex}{$k_{10}$}\right.+{\sigma }_{f\left(p\right)}^2 }}$$

(3)

where ${\sigma }_{f\left(p\right)}^2$ and ${\sigma }_e^2$ are the variances of family-within-population and the residual, respectively, and $k_{10}$ = 14.953. Nyquist [28] provided the formula for estimating the standard errors of heritabilities.

Genetic correlation (r_g) between traits was calculated following Falconer and Mackay [27]:

$$r_g = {\displaystyle \frac{{COV}_f(x,y)}{\sqrt{{\sigma }_{fx}^2 {\sigma }_{fy}^2}}}$$

(4)

where ${COV}_f$(x,y) represents the family covariance between traits X and Y, and ${\sigma }_{fx}^2$ and ${\sigma }_{fy}^2$ are their corresponding family (i.e., family-within-population) variances. The standard error of genetic correlation was estimated following Robertson [29].

2.3.2. Physiological Traits

The analysis of variance and covariance for physiological traits was conducted using the following linear model:

Y_ijkl = µ + R_i + P_j + RP_ij + F_k(P_j) + L_l(FP_jk) + ε_ijkl

(5)

where

• µ—grand mean;
• R_i—i^th replication effect, i = 1–5;
• P_j—j^th population effect, j = 1–6;
• RP_ij—effect of replication-by-population interaction;
• F_k(P_j)—k^th family effect within the j^th population, k = 1, 2, …, n (n ranges from 13 to 27);
• L_l(FP_jk)—l^th leaf effect within the k^th family within the j^th population, l = 1–2;
• ε_ijkl—residual error.

All effects in the model were assumed to be random. Expected mean squares (EMS) and expected mean cross products (EMCPs) were estimated using the SAS Varcomp procedure [25]. Significance tests of the model effects followed Satterthwaite’s approximate test procedure [26].

Similar to growth and morphology, the narrow-sense heritabilities for individuals and families were computed as follows:

$$\mathrm{Individual} \mathrm{heritability}, h_i^2= {\displaystyle \frac{3 x {\sigma }_{f\left(p\right)}^2}{{\sigma }_{e }^2+ {\sigma }_{l\left(fp\right)}^2+ {\sigma }_{f\left(p\right)}^2}}$$

(6)

$$\mathrm{Family} \mathrm{heritability}, h_f^2={\displaystyle \frac{{\sigma }_{f\left(p\right)}^2}{{\sigma }_{e }^2/k_{14}+{k_{13}\sigma }_{l\left(fp\right)}^2/k_{14}+{\sigma }_{f\left(p\right)}^2}}$$

(7)

where ${\sigma }_{f\left(p\right)}^2$, ${\sigma }_{l\left(fp\right)}^2$, and${\sigma }_{e }^2$ are the variances of family-within-population, leaf within family-within-population, and the residual, respectively, and $k_{13}$= 4.863 and $k_{14}$ = 9.719. Nyquist [28] provided the formula for estimating the standard errors of heritabilities.

The genetic correlation (r_g) between traits was calculated using Equation (4), and the standard error was determined according to Robertson [29].

3. Results

3.1. Seedling Survival

Seedling survival of P. macrocarpus after a 30-week growth period was high across all populations, averaging 95% with a range of 92–97% (

Table 2. Percentage of seedling survival after the 30-week growth period.

Population	No. of Families	Survival (%)	Range Among Families Survival (%)
1	15	95.8	75.0–100
2	14	97.3	93.8–100
3	13	93.8	75.0–100
4	21	92.3	75.0–100
5	22	94.6	75.0–100
6	27	96.5	87.5–100
Overall	112	95.0 a	75.0–100

^a Averaged from family survival (%) after 30-week growth period. There was no significant difference among populations for survival (%) after the 30-week growth period.

).

At the family level, survival rates ranged from 75% to 100%. However, no statistically significant differences in survival were observed among populations (p > 0.05), indicating that uniform nursery conditions and optimal management factors had a greater influence than genetic variation on early survival outcomes.

3.2. Growth and Morphology

Height accelerated between weeks 6 and 18 before gradually declining towards the end of the 30-week experiment (Figure 2). Population 5 achieved the highest final mean height (40.9 cm) exceeding the grand mean by 13%, while Population 1 had the lowest mean height (29.8 cm), which was 17% below the grand mean. The 27% height difference between Population 1 and Population 5 persisted despite both originating from mixed deciduous forests, suggesting potential geographic or genetic differences.

Diameters stayed fairly consistent throughout the study period, with slightly higher rates observed in younger seedlings (Figure 2). Similar to height, Population 5 performed better, with a final average diameter of 11.0 mm (6% above the overall mean of 10.4 mm). In comparison, Population 1 had an average diameter of 9.76 mm (6% below the overall mean). The 11% difference between populations aligns with the trends observed in height.

Biomass partitioning showed a preference for allocating biomass to shoots over roots, with shoot biomass averaging 6.28 g (overall mean) compared to 5.09 g for roots (Figure 3). This results in a shoot-to-root ratio (S:R) of 1.35. Shoot biomass was split almost evenly between leaves (3.03 g) and stems (3.25 g). Within the root system, taproots accounted for the majority (73.6%) of total root biomass (3.75 g), compared to fibrous roots (1.36 g). Biomass traits, except for specific leaf weight (SLWT), generally showed CV values two to three times higher than those for height or diameter, indicating greater variability. SLWT, however, showed very little variation. Population 5 yielded the highest total dry weight (14.4 g), which was 26% above the grand mean of 11.4 g, in contrast to Population 1 (8.74 g), which was 23% below the grand mean. Similar trends were consistently observed across individual biomass components.

Analysis of variance (ANOVA) for height revealed highly significant effects of population and family-within-population (p < 0.01) across all height measurements (

Table 3. Components of variance (percentage of total variance) and significant tests for heights of P. macrocarpus seedlings grown in the nursery.

Source of Variation	df	H3	H6	H9	H12	H15	H18	H21	H24	H27	H30
Replication	15	0.02 ns	0.18 ns	0.11 ns	1.19 **	4.57 **	7.72 **	10.6 **	9.45 **	10.0 **	9.89 **
Population	5	11.8 **	9.40 **	15.2 **	17.6 **	11.6 **	7.74 **	8.9 **	9.67 **	10.6 **	11.4 **
Replication × Population	75	0.00 ns	0.00 ns	0.00 ns	0.20 ns	0.00 ns	0.00 ns	0.00 ns	0.00 ns	0.13 ns	0.00 ns
Family (Population)	106	24.1 **	25.4 **	30.7 **	28.1 **	21.6 **	15.1 **	13.4 **	13.6 **	13.5 **	12.8 **
Residual	1479	64.0	65.1	54.1	52.9	62.3	69.4	67.1	67.3	65.7	65.9

ns: not significant; **: significant at p < 0.01.

). Replication effects became significant after week 12 (p < 0.01), although replication-by-population interactions remained nonsignificant. Family-within-population variance was dominant in early growth, contributing up to 32% of the total variability, but decreased to 13% by week 30. Meanwhile, residual variance consistently represented the largest proportion, ranging from 53% to 69% of the total variability.

Similar patterns emerged in diameter growth (

Table 4. Components of variance (percentage of total variance) and significant tests for diameters of P. macrocarpus seedlings grown in the nursery.

Source of Variation	df	D12	D15	D18	D21	D24	D27	D30
Replication	15	0.16 ns	0.00 ns	1.47 **	3.16 **	1.79 **	4.35**	4.77 **
Population	5	2.72 *	3.27 **	3.34 **	2.57 *	2.61 *	4.28**	4.67 **
Replication × Population	75	0.78 ns	0.55 ns	1.43 *	0.57 ns	0.58 ns	0.49 ns	0.03 ns
Family (Population)	106	18.1 **	19.4 **	18.6 **	21.0 **	18.0 **	16.5 **	17.0 **
Residual	1479	78.3	76.8	75.1	72.7	77.0	74.4	73.5

ns: not significant; *: significant at p < 0.05; **: significant at p < 0.01.

), with family-within-population effects explaining 16–21% of the variance, while residuals accounted for 72–78%.

ANOVA results for biomass traits (

Table 5. Components of variance (percentage of total variance) and significant tests for biomass of P. macrocarpus seedlings grown in the nursery.

Source of Variation	df	LEAF	STEM	TROOT	FROOT	TOTAL	SHOOT	ROOT	S:R	SLWT
Replication	15	7.79 **	10.1 **	6.14 **	4.49 **	9.49 **	9.79 **	6.44 **	1.20 **	12.9 **
Population	5	15.2 **	10.2 **	7.59 **	4.46 **	11.0 **	13.2 **	7.09 **	8.04 **	2.28 **
Replication × Population	75	0.18 ns	0.00 ns	0.08 ns	0.00 ns	0.21 ns	0.10 ns	0.00 ns	0.00 ns	0.64 ns
Family (Population)	106	10.3 **	14.8 **	13.3 **	11.6 **	13.7 **	12.9 **	3.26 **	8.72 **	2.62 **
Residual	1479	66.5	64.8	72.9	79.5	65.7	64.06	73.2	82.0	81.6

ns: not significant; **: significant at p < 0.01.

) showed significant effects of replication, population, and family (p < 0.01) for most traits, except for SLWT, where replication effects explained five times more variance than genetic components.

Heritability estimates for height (Figure 4) were moderate to high, ranging from 0.49 to 1.00 for individuals, and from 0.74 to 0.90 for families. A decline in height heritabilities was observed with increasing seedling age, around H12.

Diameter heritabilities (Figure 4) were similarly high, with individual estimates ranging from 0.55 to 0.68 and family estimates from 0.77 to 0.81, with family estimates consistently exceeding individual values by approximately 32%.

Biomass traits had moderate heritabilities (Figure 5), with individual estimates ranging from 0.39 to 0.56 and family estimates ranging from 0.69 to 0.77. Exceptions included S:R (0.291 individuals; 0.61 families) and SLWT (0.09 individuals; 0.32 families), which showed lower genetic control.

Strong genetic correlations, ranging from 0.39 to 1.00, were observed among heights at successive ages (

Table 6. Estimates of genetic correlations (above diagonal) and their standard errors (below diagonal) among the heights of P. macrocarpus seedlings grown in the nursery.

H3	H6	H9	H12	H15	H18	H21	H24	H27	H30
H3	****	0.66	0.66	0.72	0.67	0.55	0.52	0.50	0.42	0.39
H6	0.07	****	0.88	0.81	0.79	0.70	0.70	0.69	0.64	0.60
H9	0.07	0.03	****	0.96	0.92	0.81	0.80	0.79	0.69	0.66
H12	0.06	0.04	0.01	****	0.96	0.86	0.85	0.83	0.73	0.70
H15	0.07	0.05	0.03	0.01	****	0.95	0.92	0.91	0.82	0.79
H18	0.09	0.07	0.05	0.04	0.02	****	0.98	0.96	0.92	0.89
H21	0.09	0.07	0.05	0.04	0.03	0.01	****	1.00	0.98	0.96
H24	0.10	0.07	0.05	0.04	0.03	0.01	0.00	****	0.99	0.98
H27	0.10	0.08	0.07	0.06	0.05	0.03	0.01	0.01	****	1.00
H30	0.11	0.08	0.07	0.07	0.05	0.03	0.02	0.01	0.00	****

). However, these correlations tended to decrease over longer time intervals.

Diameter correlations across different ages (

Table 7. Estimates of genetic correlations (above diagonal) and their standard errors (below diagonal) among the diameters of P. macrocarpus seedlings grown in the nursery.

	D12	D15	D18	D21	D24	D27	D30
D12	****	0.97	0.91	0.88	0.89	0.87	0.86
D15	0.02	****	0.99	0.97	0.96	0.95	0.91
D18	0.03	0.01	****	0.98	0.98	0.98	0.95
D21	0.04	0.02	0.01	****	1.00	0.99	0.97
D24	0.04	0.02	0.02	0.01	****	0.99	0.97
D27	0.05	0.03	0.02	0.01	0.01	****	1.00
D30	0.05	0.03	0.02	0.01	0.01	0.00	****

) were near-perfect, ranging from 0.86 (±0.05) to 1.00 (±0.01).

Genetic correlations between height and diameter ranged from 0.55 to 0.92 (

Table 8. Estimates of genetic correlations (r_g) and their standard errors (s.e.) between height and diameter and biomass traits of P. macrocarpus seedlings grown in the nursery.

rg	H3	H6	H9	H12	H15	H18	H21	H24	H27	H30
D12	0.55	0.59	0.80	0.79	0.73	0.62	0.64	0.65	0.58	0.56
s.e	0.09	0.08	0.05	0.05	0.06	0.08	0.08	0.08	0.09	0.09
D15	0.65	0.68	0.86	0.89	0.86	0.77	0.76	0.76	0.70	0.69
s.e	0.08	0.07	0.04	0.04	0.04	0.06	0.06	0.06	0.07	0.08
D18	0.68	0.72	0.87	0.92	0.88	0.75	0.73	0.74	0.66	0.64
s.e	0.07	0.07	0.04	0.03	0.04	0.06	0.07	0.07	0.08	0.08
D21	0.62	0.72	0.81	0.87	0.87	0.76	0.73	0.73	0.65	0.62
s.e	0.08	0.07	0.05	0.04	0.04	0.06	0.07	0.07	0.08	0.08
D24	0.60	0.70	0.82	0.87	0.88	0.78	0.74	0.75	0.68	0.66
s.e	0.08	0.07	0.05	0.04	0.04	0.06	0.07	0.07	0.08	0.08
D27	0.57	0.70	0.82	0.87	0.88	0.81	0.78	0.78	0.72	0.70
s.e	0.09	0.07	0.05	0.04	0.04	0.06	0.06	0.06	0.07	0.07
D30	0.61	0.74	0.83	0.87	0.87	0.80	0.77	0.77	0.70	0.68
s.e	0.08	0.07	0.05	0.04	0.04	0.06	0.06	0.06	0.07	0.07
LEAF	0.30	0.58	0.53	0.55	0.66	0.75	0.78	0.79	0.82	0.81
s.e	0.12	0.09	0.09	0.09	0.08	0.07	0.06	0.06	0.05	0.05
STEM	0.54	0.74	0.79	0.82	0.88	0.89	0.88	0.88	0.86	0.84
s.e	0.09	0.06	0.05	0.05	0.04	0.03	0.04	0.03	0.04	0.04
TROOT	0.50	0.50	0.53	0.57	0.67	0.73	0.63	0.59	0.58	0.56
s.e	0.10	0.10	0.09	0.09	0.07	0.07	0.08	0.09	0.09	0.09
FROOT	0.40	0.56	0.63	0.69	0.78	0.81	0.80	0.79	0.75	0.74
s.e	0.11	0.10	0.08	0.08	0.06	0.06	0.06	0.06	0.07	0.07
TOTAL	0.50	0.67	0.69	0.73	0.82	0.88	0.84	0.83	0.82	0.80
s.e	0.09	0.08	0.07	0.06	0.05	0.04	0.04	0.05	0.05	0.05
SHOOT	0.45	0.70	0.71	0.73	0.82	0.86	0.87	0.87	0.87	0.86
s.e	0.10	0.07	0.07	0.06	0.05	0.04	0.04	0.04	0.03	0.04
ROOT	0.51	0.55	0.59	0.64	0.75	0.80	0.72	0.68	0.67	0.65
s.e	0.10	0.09	0.08	0.08	0.06	0.05	0.07	0.07	0.08	0.08
S:R	−0.12	0.21	0.16	0.12	0.08	0.07	0.22	0.26	0.31	0.32
s.e	0.13	0.13	0.14	0.13	0.13	0.14	0.14	0.14	0.13	0.13
SLWT	−0.14	0.07	0.17	0.18	0.22	0.11	0.11	0.10	0.04	0.05
s.e	0.18	0.18	0.18	0.18	0.18	0.19	0.19	0.19	0.20	0.20

). Height–biomass correlations (Table 8) strengthened with age, reaching a peak of 0.89 for stem biomass.

Diameter–biomass correlations (

Table 9. Estimates of genetic correlations (r_g) and their standard errors (s.e.) between diameters and biomass traits of P. macrocarpus seedlings grown in the nursery.

rg	D12	D15	D18	D21	D24	D27	D30
LEAF	0.49	0.62	0.57	0.58	0.56	0.60	0.59
s.e.	0.10	0.09	0.09	0.09	0.09	0.09	0.09
STEM	0.73	0.84	0.85	0.85	0.85	0.89	0.88
s.e.	0.07	0.05	0.04	0.04	0.04	0.03	0.03
TROOT	0.44	0.67	0.61	0.59	0.54	0.54	0.51
s.e.	0.11	0.08	0.09	0.09	0.09	0.09	0.10
FROOT	0.62	0.77	0.71	0.74	0.74	0.73	0.72
s.e.	0.09	0.07	0.07	0.07	0.07	0.07	0.07
TOTAL	0.62	0.80	0.76	0.76	0.74	0.76	0.74
s.e.	0.08	0.05	0.06	0.06	0.06	0.06	0.06
SHOOT	0.64	0.78	0.76	0.76	0.76	0.80	0.78
s.e.	0.08	0.06	0.06	0.06	0.06	0.05	0.05
ROOT	0.52	0.74	0.68	0.67	0.63	0.63	0.60
s.e.	0.10	0.07	0.08	0.08	0.08	0.08	0.09
S:R	0.20	0.07	0.13	0.15	0.21	0.27	0.28
s.e.	0.14	0.14	0.14	0.13	0.14	0.13	0.13
SLWT	0.23	0.10	0.01	0.13	0.18	0.08	0.09
s.e.	0.19	0.19	0.19	0.19	0.19	0.19	0.19

) showed a similar trend, ranging from 0.44 to 0.89.

Biomass traits (

Table 10. Estimates of genetic correlations (above diagonal) and their standard errors (below diagonal) among biomass traits of P. macrocarpus seedlings grown in the nursery.

	LEAF	STEM	TROOT	FROOT	TOTAL	SHOOT	ROOT	S:R	SLWT
LEAF	****	0.85	0.67	0.82	0.91	0.95	0.76	0.30	−0.09
STEM	0.04	****	0.71	0.87	0.94	0.97	0.80	0.24	0.02
TROOT	0.08	0.07	****	0.66	0.87	0.72	0.98	−0.40	−0.31
FROOT	0.06	0.04	0.08	****	0.89	0.88	0.80	0.03	0.04
TOTAL	0.03	0.02	0.03	0.03	****	0.96	0.94	0.02	−0.13
SHOOT	0.01	0.01	0.07	0.04	0.01	****	0.81	0.27	−0.03
ROOT	0.06	0.05	0.01	0.05	0.02	0.05	****	−0.31	−0.24
S:R	0.14	0.13	0.12	0.15	0.14	0.13	0.13	****	0.46
SLWT	0.21	0.19	0.20	0.21	0.20	0.20	0.20	0.22	****

) generally showed high inter-trait correlations (0.66–0.98), indicating integrated developmental regulation among these traits, except for S:R and SLWT, which had weak or negative correlations with root components. Notably, shoot biomass traits (leaf and stem) showed stronger integration compared to root traits (taproot and fibrous root).

3.3. Physiological Traits

Net photosynthesis (A), transpiration (E), and water-use efficiency (WUE) of P. macrocarpus seedlings (

Table 11. Population means and standard deviation (SD), range of family means, grand means, and coefficients of variation (C.V.s) for net photosynthesis (A), transpiration (E), and water-use efficiency (WUE) of P. macrocarpus seedlings grown in the nursery.

Population
	1	2	3	4	5	6
Trait	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Grand Mean	C.V.a (%)
A b	7.98 (3.54)	7.93 (3.09)	7.88 (4.18)	8.54 (3.17)	9.08 (3.67)	8.40 (3.32)	8.39	41.7
Range	5.66–10.0	5.99–11.6	5.31–10.7	6.29–11.7	6.12–11.0	5.15–11.1
E c	1.49 (0.76)	1.39 (0.65)	1.31 (0.73)	1.41 (0.62)	1.35 (0.63)	1.44 (0.74)	1.40	49.2
Range	1.13–2.18	1.09–1.96	0.89–1.86	0.99–2.03	0.98–1.90	0.91–2.48
WUE d	5.82 (2.35)	6.26 (2.23)	6.45 (3.01)	6.58 (2.46)	7.28 (2.51)	6.54 (2.58)	6.56	39.1
Range	4.32–7.78	5.34–7.20	4.95–8.40	5.28–9.31	6.03–8.86	4.86–8.20

^a Coefficient of variation (%) based on individual observations. ^b Measurement unit: μmol m⁻² s⁻¹. ^c Measurement unit: mmol m⁻² s⁻¹. ^d Measurement unit: μmol CO₂ /mmol H₂O.

) averaged 8.39 µmol m⁻² s⁻¹, 1.4 mmol m⁻² s⁻¹, and 6.56 µmol CO₂/mmol H₂O had notable variability among seedlings. Population 5 again demonstrated the highest mean net photosynthesis (9.08 µmol m⁻² s⁻¹), an 8% increase above the grand mean, consistent with its superior growth and biomass performance. While Population 1 did not have the lowest net photosynthesis (7.98 µmol m⁻² s⁻¹), it recorded the highest transpiration rate (1.49 mmol m⁻² s⁻¹), contrasting with Population 3, which had the lowest E (1.31 mmol m⁻² s⁻¹). Water-use efficiency varied significantly among populations, with Population 5 achieving the highest WUE (7.28 µmol CO₂/mmol H₂O, 11% above the grand mean) due to its combination of high A and moderate E. Conversely, Population 1, characterised by lower A and higher E, exhibited the lowest WUE (5.82 µmol CO₂/mmol H₂O), a 20% difference that underscores distinct water conservation strategies among geographically diverse populations.

Analysis of variance for physiological traits (

Table 12. Components of variance (percentage of total variance) and significant tests for physiological traits of P. macrocarpus seedlings grown in the nursery ^a.

Source of Variation	df	A	E	WUE
Replication	4	0.27 ns	3.90 **	14.6 **
Population	5	0.00 ns	0.00 ns	2.34 **
Replication × Population	20	3.58 **	1.44 *	0.98 ns
Family (Population)	106	12.7 **	13.4 **	6.94 **
Leaf (Family Population)	112	0.00 ns	0.00 ns	0.00 ns
Residual	842	83.5	81.3	75.1

^a Only 5 replications were assessed for physiological traits, and two leaves from each seedling were measured. ns: not significant; *: significant at p < 0.05; **: significant at p < 0.01.

) revealed highly significant family effects (p < 0.01) for all physiological traits (A, E, and WUE), indicating substantial genetic influences on their phenotypic variation. Population effects were significant only for WUE (p < 0.01), while replication effects were significant for E and WUE (p < 0.01). Significant replication-by-population interactions were found for A (p < 0.01) and E (p < 0.05), suggesting environmental modulation of these traits. Residual variance accounted for the majority of total variability (75–83%), reflecting the trends observed in growth and biomass traits.

Heritability estimates for physiological traits (Figure 5) were low to moderate: individual heritabilities ranged from 0.26 (±0.08) for WUE to 0.43 (±0.09) for E, while family heritabilities varied between 0.47 (±0.10) and 0.62 (±0.07). Transpiration showed the highest heritability among physiological traits. The standard errors for these estimates (11–29%) were larger than those for growth traits, indicating greater environmental influence on physiological responses.

Net photosynthesis and transpiration showed a strong positive genetic correlation (0.78 ± 0.12). However, physiological traits demonstrated only weak or no correlation with growth and biomass parameters (

Table 13. Estimates of genetic correlations (r_g) and their standard errors (in parentheses) among physiological traits and growth and biomass traits of P. macrocarpus seedlings grown in the nursery ^a.

rg	A	E	WUE
H3	0.07 (0.14)	−0.02 (0.14)	0.16 (0.16)
H6	0.05 (0.14)	−0.05 (0.14)	0.15 (0.16)
H9	−0.02 (0.14)	−0.11 (0.13)	0.19 (0.15)
H12	0.04 (0.14)	−0.05 (0.13)	0.17 (0.15)
H15	0.06 (0.14)	−0.07 (0.14)	0.19 (0.16)
H18	0.06 (0.15)	−0.02 (0.14)	0.09 (0.17)
H21	0.07 (0.15)	−0.08 (0.14)	0.24 (0.16)
H24	0.10 (0.15)	−0.06 (0.14)	0.26 (0.16)
H27	0.09 (0.15)	−0.12 (0.14)	0.31 (0.16)
H30	0.07 (0.15)	−0.15 (0.15)	0.32 (0.16)
D12	−0.05 (0.14)	−0.13 (0.14)	0.13 (0.16)
D15	−0.08 (0.15)	−0.16 (0.14)	0.14 (0.16)
D18	−0.03 (0.15)	−0.09 (0.14)	0.09 (0.16)
D21	−0.01 (0.14)	−0.100 (0.14)	0.04 (0.16)
D24	−0.02 (0.14)	−0.09 (0.14)	0.07 (0.16)
D27	−0.04 (0.14)	−0.14 (0.14)	0.11 (0.16)
D30	−0.02 (0.14)	−0.14 (0.14)	0.14 (0.16)
LEAF	−0.17 (0.15)	−0.32 (0.14)	0.27 (0.17)
STEM	0.01 (0.15)	−0.14 (0.14)	0.22 (0.16)
TROOT	0.01 (0.15)	−0.13 (0.14)	0.20 (0.16)
FROOT	−0.08 (0.15)	−0.14 (0.15)	0.03 (0.17)
TOTAL	−0.06 (0.15)	−0.22 (0.14)	0.24 (0.16)
SHOOT	−0.07 (0.15)	−0.23 (0.14)	0.26 (0.16)
ROOT	−0.02 (0.15)	−0.15 (0.14)	0.17 (0.17)
S:R	−0.13 (0.15)	−0.19 (0.14)	0.12 (0.16)
SLWT	0.14 (0.15)	0.06 (0.15)	0.06 (0.18)

^a Only five replications were included in the analysis.

), limiting their usefulness as proxies for long-term productivity. WUE correlations with its components (A, E) were not included due to inherent formulaic dependencies. WUE exhibited marginally stronger associations with growth traits (e.g., height, stem biomass) than A or E, although these correlations remained nonsignificant (r < 0.3) with large standard errors. These findings suggest that physiological processes such as photosynthesis and transpiration may operate independently of morphological growth under the studied conditions, possibly due to differing resource allocation strategies.

4. Discussion

4.1. Nursery Performance and Early Growth Dynamics

The high survival rates of P. macrocarpus seedlings (average: 95%; range: 75–100%) in nursery conditions demonstrate their adaptability to controlled environments (Table 2). This strong survival rate aligns with findings in other tropical hardwoods, where optimal resource availability in nurseries reduces abiotic stress, supporting healthy establishment and early growth [30,31]. Such controlled settings are essential for detecting genetic variation in juvenile traits, which is vital for breeding programs targeting species with long life cycles.

4.2. Growth and Morphological Traits

The observed increase in height and diameter (Figure 2) and biomass (Figure 3) after 30 weeks demonstrates significant early growth. The rapid height gain between weeks 6 and 18 aligns with typical exponential growth phases in juvenile trees. The decline observed after transplanting (week 12) is likely due to transplant shock and the redirection of resources towards root development [32,33]. A further slowdown after week 24 coincided with the species’ seasonal dormancy (January–March), during which P. macrocarpus sheds foliage and reallocates resources to structural and storage tissues. In contrast, diameter growth remained steady throughout the season (Figure 2), reflecting patterns seen in species such as Pinus taeda [34] and Douglas-fir [35], where radial growth continues even when height growth diminishes. This varied response indicates that apical and radial meristems react differently to dormancy cues. Reductions in diameter growth at later stages may also suggest increased competition among seedlings at higher densities [36,37].

The shift favouring shoots (S:R = 1.35) and the dominance of taproots (3.75 g) over fibrous roots (1.36 g) (Figure 3) align with strategies seen in deciduous species. Taproots are crucial for improving drought resilience by reaching deeper into the soil during dormancy [38,39]. Meanwhile, reduced investment in fibrous roots, which become less active in dry seasons, helps optimise survival during seasonal water shortages.

Significant differences at the population level were evident, with Population 5 consistently exceeding Population 1 in height (27%), diameter (11%), and total biomass (39%) (Figure 2 and Figure 3). These differences probably result from local adaptation, as Population 5 came from an environment similar to the nursery, whereas Population 1 was from native conditions that varied. This environmental filtering aligns with the variability observed in germination responses in P. macrocarpus [40] and underscores the necessity for provenance-specific conservation strategies.

ANOVA confirmed significant genetic variation among both populations and families (Table 3, Table 4 and Table 5), consistent with findings in other outcrossing tropical tree species [41,42]. The dominance of within-family variation aligns with expectations for insect-pollinated species like P. macrocarpus, where high pollen diversity promotes heterozygosity [23]. Population differentiation in this study exceeded what is usually observed in conifers [43,44], likely because fragmented habitats limit gene flow—a situation that is even more pronounced in insect-pollinated taxa with limited pollinator mobility [45,46]. The significant family-within-population effect further supports previous isozyme results, which indicate that most genetic variation in P. macrocarpus occurs within populations [17]. The random selection of seed trees, without bias towards superior phenotypes, may have contributed to maintaining this genetic diversity [47].

Family-level variance for height peaked early and then declined to 13% by week 30 (Table 3), indicating stronger genetic expression under low competition conditions [48]. This decline in family effects with age reflects the competition-driven suppression of genetic signals observed in species such as Picea abies [48], Pinus taeda [49], and Pseudotsuga menziesii [50]. Maternal effects, influenced by seed weight, significantly affected early growth (Figure 6), with height correlations decreasing from 0.73 at week 3 to 0.38 at week 30. The diminishing influence over time aligns with the general pattern of maternal provisioning impacts [51,52].

Moderate to high heritability estimates for height (individual: 0.49–1.00; family: 0.74–0.90) and diameter (individual: 0.559–0.68; family: 0.77–0.81) (Figure 4) surpass those reported for many tropical species [53,54], probably because controlled nursery conditions reduce environmental noise. The decrease in height heritability with age (Figure 4) matches increased competition that masks genetic differences [36,37], while the consistent heritability of diameter (Figure 4) suggests stronger maternal or pleiotropic effects.

Biomass traits showed moderate heritability (e.g., total biomass: individual 0.52; family 0.76), except for S:R (individual 0.29) and SLWT (individual 0.09) (Figure 5). While S:R’s plasticity reflects adaptive trade-offs between shoot and root investment under varying resource conditions [55,56], SLWT’s limited value for selection stems from its sensitivity to microclimatic fluctuations [57,58].

Strong genetic correlations among height, diameter, and biomass traits (Table 6, Table 7, Table 8, Table 9 and Table 10) support the notion of pleiotropic gene action, allowing for the simultaneous selection of multiple desirable traits. Height–diameter correlations (0.55–0.92) and strong biomass links (e.g., LEAF-STEM: 0.85) suggest coordinated developmental regulation. Weak correlations for S:R and SLWT with other traits highlight their environmental sensitivity, requiring cautious interpretation [57,58]. Early-age correlations (e.g., H3-LEAF: 0.30) emphasise the temporary influence of maternal effects, while the stability of late-age correlations (e.g., H30-TOTAL: 0.80) indicates reliable selection after dormancy [32,59]. The potential for early culling (e.g., removing low-performing families at week 12) supports cost-effective nursery management strategies [22,60]. However, field validation remains crucial to account for genotype-by-environment (G × E) interactions.

4.3. Physiological Traits

The net photosynthetic rate (A) in P. macrocarpus seedlings, which ranges from 7.88 to 9.08 µmol m⁻² s⁻¹ across populations (Table 11), shows Population 5 (northern Thailand) reaching the highest values. Although differences among populations are not statistically significant for A (Table 12), the superior performance of Population 5 in nursery conditions—closely resembling its native climate—indicates local adaptation. This phenomenon is well-documented in tree species [36,61,62], reflecting their acclimation to local temperature, humidity, and light conditions [62,63].

Transpiration rate (E) showed similar patterns, with Population 1 (western Thailand) exhibiting the highest E (1.49 mmol m⁻² s⁻¹) despite originating from a cooler native habitat (Table 11). Higher transpiration in warmer environments supports findings in Eucalyptus camaldulensis, where heat stress is linked to increased stomatal conductance [64]. This observed plasticity highlights the species’ ability to adapt its water-use strategies to new conditions, although such responses may involve trade-offs in water-limited environments.

Water-use efficiency (WUE), calculated as A/E, was highest in Population 5 (7.28 µmol CO₂ mmol⁻¹ H₂O) and lowest in Population 1 (5.82 µmol CO₂ mmol⁻¹ H₂O) (Table 11). The combination of high A and moderate E in Population 5 reflects adaptations seen in xeric Populus trichocarpa clones that optimise carbon gain during drought [65]. Conversely, Population 1′s low WUE, caused by high E and modest A, likely exacerbated water deficits under containerised nursery conditions, leading to reduced biomass accumulation. These findings highlight the complex interaction between physiological adaptation and microenvironmental constraints in influencing seedling performance.

Significant family effects within populations for A, E, and WUE (Table 12) highlight considerable genetic variation in gas exchange traits, aligning with patterns in growth and biomass. However, differences at the population level were only significant for WUE, indicating that selective pressures on water-use efficiency may be more influential at larger geographic scales. Similar research in Picea abies and Alnus rubra reveals inconsistent genetic structuring of gas exchange traits, often attributed to varying environmental gradients and gene flow patterns [66,67]. The minimal population effects on A and E contrast with findings in Larix occidentalis, where provenance differences in photosynthesis were absent [68]. This discrepancy could be attributed to the uniform influence of nursery conditions or limited sampling of extreme ecotypes. Additionally, significant interactions between replication and population suggest that specific environmental contexts and microenvironmental heterogeneity, such as light availability or irrigation schedules, influence genetic differences in physiology.

Narrow-sense heritability estimates for A (0.40 ± 0.09), E (0.43 ± 0.09), and WUE (0.26 ± 0.08) were moderate, with low standard errors confirming genetic control (Figure 5). These values agree with heritabilities reported for crops where photosynthetic traits display similar environmental sensitivity [69,70]. The lower heritability of WUE compared to E emphasises its dual reliance on stomatal regulation and carboxylation efficiency, both of which are highly responsive to diurnal and seasonal changes [71]. Ontogenetic shifts further complicate the expression of physiological traits; declining gas exchange rates with seedling age, as seen in Prunus serotina [72] and Robinia pseudoacacia [73], reflect metabolic reallocations from leaf production to structural biomass growth. Leaf morphological plasticity, documented in P. macrocarpus (Liengsiri, pers. obs.), introduces additional variability as SLWT and longevity affect photosynthetic capacity [66,73]. Despite these challenges, the moderate heritabilities suggest potential for improving physiological traits through selective breeding, especially under stable nursery conditions.

The strong genetic correlation between A and E (0.78 ± 0.12) suggests coordinated stomatal and biochemical control of gas exchange, a pattern also seen in Alnus rubra [67]. However, the weak correlations between physiological traits and growth or biomass parameters (Table 13) limit their usefulness as direct indicators of long-term productivity. While Pseudotsuga menziesii exhibits strong links between WUE and growth under drought conditions [74], such relationships are inconsistent in P. macrocarpus, likely due to differences in methodology and timing. Gas exchange measurements reflect momentary leaf-level processes, whereas growth measures cumulative physiological performance across tissues and over time [7,9]. Environmental variation further separates leaf-level physiology from whole-plant outcomes, supporting critiques of the limited ability of single-point measurements to predict field performance [19,66].

5. Conclusions

This study confirms significant genetic variation and heritable control affecting growth, biomass accumulation, and physiological traits in P. macrocarpus seedlings. It provides essential foundational insights into the genetic makeup of seedling performance, supporting the species’ potential for targeted genetic improvement and early selection in breeding programs. However, the temporal and environmental limitations of this research highlight the need to extend evaluations to mature growth stages and field conditions. Such long-term studies are crucial for understanding age-related genetic changes, shifts in trait heritability over development, and the impact of environmental stressors on phenotypic expression.

Given P. macrocarpus’s broad native distribution and documented population differentiation across various landscapes [17], future trials must carefully implement multi-site experimental designs to evaluate genotype-by-environment (G×E) interactions. These data are essential for establishing breeding zones, defining conservation units, and tailoring deployment strategies to match genetic resources with specific ecological conditions of each site. Moreover, the species’ deciduous phenology and interannual leaf morphological plasticity (Liengsiri, personal observation) require repeated, multi-season assessments under different environmental regimes to produce reliable estimates of physiological trait stability.

Among these traits, water-use efficiency (WUE) stands out as a particularly promising candidate for selective breeding. Although physiological measurements are labour-intensive, incorporating WUE into selection criteria could significantly improve seedling resilience in drought-prone or degraded areas, offering practical benefits for reforestation and ex situ conservation projects. In Thailand, where extensive deforestation has led to land degradation characterised by decreased soil fertility, reduced hydrological retention, and notable microclimatic fluctuations, establishing seedlings remains a significant challenge. While silvicultural practices can enhance early survival, their economic costs often limit large-scale efforts. Focusing on genotypes with higher WUE provides a scalable, cost-effective way to increase seedling productivity in water-scarce conditions, supported by evidence that WUE is a key factor in plant performance in arid environments [75,76].

Notably, this study also highlights potential effects of inbreeding on growth performance. Families with lower outcrossing rates showed significantly slower growth (Table 1), consistent with previous findings of inbreeding depression in P. macrocarpus populations. This depression, which is well documented in forest trees, may bias estimates of additive genetic variance and heritability. To minimise these effects, breeding programs should adopt strategies that explicitly limit inbreeding, such as controlled crossing schemes or maintaining diverse founder populations, to ensure accurate selection and sustain long-term genetic gains.

In conclusion, although these findings confirm the genetic potential of P. macrocarpus for trait improvement, the success of long-term breeding and conservation efforts depends on ongoing research in three key areas: (1) spatiotemporal patterns of genetic diversity across the species’ range, (2) the extent and factors driving G×E interactions within different landscapes, and (3) the physiological mechanisms that enable drought adaptation and resilience. Focusing on these areas will support the creation of evidence-based strategies for germplasm conservation, targeted breeding, and adaptive silviculture, ultimately boosting the ecological and economic value of P. macrocarpus amid increasing climate uncertainties.