Influence of stand density, site, age, and competition on the timber assortment structure of Chinese fir plantations Authors: Xiaoyan Li, Aiguo Duan, Jianguo Zhang Published: 2024 ABSTRACT The growth of timber assortment is related to site conditions and forest management measures. To our knowledge, however, the effects and relative importance of these factors on the growth of different timber assortments of Chinese fir have not yet been explored. Based on data from long-term fixed observation plots in the northwest of the Chinese fir distribution zone, we examined the effects of planting density and the site index on the timber assortment structure of Chinese fir. Furthermore, we quantified the relative importance of stand factors (age, planting density, stand density index, number of living trees, Gini coefficient, and site index) in determining the growth of various timber assortments. The results showed that the wood yield of all timber assortment types increased as the site quality improved. The outturn and outturn rate of large- and medium-diameter timber were negatively correlated with the planting density. Before the age of 22, planting density had a negative impact on total timber yield and small-diameter timber outturn. After 22 years, planting density had a positive impact on total timber yield and small-diameter timber outturn. SDI (stand density index), age, and SD (stand density) were the most important factors influencing total timber yield and small-diameter timber. The relative importance of age to large-diameter timber was the greatest, followed by SI (site index) and SD. The effects of age, SD, SI, and PD (planting density) on the medium-diameter timber were greater than those of SDI and Gini coefficient. The results indicate small- and medium-diameter timber as the primary target timber assortment types to be cultivated in the study area, while large-diameter timber cannot be effectively produced under the current site conditions. Our findings provide valuable insights for the management of Chinese fir plantations and the directed cultivation of target timber assortments in the northwest of the Chinese fir distribution zone. Introduction Although China has consistently maintained “double growth” in its forest area and forest stock in recent years, its forest stock is only 69% of the global average per unit area1. Large-diameter and precious wood resources have grown in importance as global strategic resources due to the rising demand for these products in many different nations2,3. However, due to an imbalance in China’s domestic supply and demand for wood, there has long been a structural shortage of large-diameter timber and rare tree species. Moreover, China’s wood is at risk of domestic shortage and import restriction because of the influence of national policy control, scarcity of global wood resources, and policy restricting log export quantity. Based on this, increasing the domestic supply of wood is fundamental. Therefore, resolving the conflict between supply and demand for wood and ensuring ecological security can be achieved by the efficient forest tree cultivation. In terms of effective cultivation technology, the main goal should be to target large- and medium-diameter timber species for directional cultivation. This will reduce the imbalance in China’s timber supply structure. Large-diameter timber cultivation has become one of the primary research areas in oriented timber tree cultivation, with scholars proposing numerous cultivation technology models for different tree species4,5. Timber assortment structure is the composition structure of different timber species in the stand6 and is commonly expressed by the outturn and outturn rate of the timber assortment. A wide range of ecological factors influence growth and yield distribution. Understanding the impact of these factors on the timber assortment structure of stands is critical. The initial planting density is an important cultivation factor influencing the plantation yield and wood quality. As a result, numerous studies have been conducted to determine the relationship between initial planting density and growth characteristics7,8, as well as wood yield9 and wood properties10. According to Zeide (2004)11, forestry is the science of optimizing stand density. The initial and final stand density predetermine the success of forestry activities, as well as the achievement of expected results in terms of wood quality and quantity. In general, neglecting the importance of planting density reduces stand productivity12. Danilov et al. (2023)12 demonstrated that the quality of spruce wood produced over a short rotation period was dependent on initial planting density and timely regulation. Liu et al. (2019)13 reported that as age increased, the proportion of medium- and large-diameter timber in stands with a low initial planting density increased faster than in stands with a high initial planting density. Thus, the initial planting density must be carefully selected to maximize growth rate and wood quality10. Stand density management has long been recognized as a key afforestation intervention, allowing forest managers to maximize growth and yield11. Thinning is an important means of controlling stand density due to its ability to effectively control stand competition, release forest growth potential, and accelerate the formation of large- and medium-diameter timber4. Sun et al. (2023)14 found that moderate thinning increased individual tree volume and the stand diameter. Allen et al. (2021)9 revealed that while thinning can provide an early return on investment and reduce densification-related mortality, optimized thinning should consider the effects of stand age, site productivity, the time when thinning began, and the frequency of thinning to correctly evaluate the relationship between growth and density. Foresters have attempted to select the stand density that maximizes total stand volume, yet no definitive answer has been determined. This is because the impact of density on growth is related to factors such as tree size and age11. Site quality reflects the stand’s productive capacity and determines its productivity. It also affects tree growth potential and competitiveness15. The site index is a widely used site quality evaluation index that has gained global recognition16. Numerous studies have been conducted to determine the relationship between the site index and forest growth, productivity15, wood yield17, wood quality, and performance18. The maximum production capacity of tree species and wood quality differ depending on the site quality, and in general, the better the site quality, the greater the tree species’ wood production potential and wood quality17. The majority of studies identify the site index as the most important factor determining the formation of large- and medium-diameter timber5,19. Forest stand structure is a fundamental feature of the forest ecosystem and includes the characteristics and relationships of living trees. Optimizing the stand structure can maximize the function benefits of the forest ecosystem20. Numerous studies have found that forest structure (e.g., stand density, age, and tree size) has an impact on stand productivity and wood yield21,22. Stand density, which includes the number of trees, stand basal area, leaf area index, and stand density index, is one of the most significant stand structural factors affecting growth. Age is another structural characteristic that markedly affects productivity and biomass. By increasing tree size and size variation, stand age can boost biomass and productivity23. The heterogeneity of stand structure is closely associated with forest growth, productivity, and wood yield24,25. The Gini coefficient is a size heterogeneity index that is frequently used to measure stand structure and establish a relationship between stand structure and productivity26,27. Although many studies have examined the relationship between stand productivity or timber yield and stand structural characteristics, no single stand structure index can predict productivity or timber yield across forest types, and changes in stand productivity or timber yield are typically explained by more than one stand structure attribute21,24. Understanding the effects of various stand factors on timber assortment structure or wood yield can help us to accurately and timely adjust the stand structure in response to local conditions, promoting the growth and development of target tree species and increasing wood yield. Furthermore, the trade-off between stand structural characteristics and productivity or timber yield varies with time, spatial scale, and afforestation management28. However, it is unclear to what extent planting density, site conditions, and stand characteristics influence timber yield in stands, particularly for trees of different diameter classes. Chinese fir (Cunninghamia lanceolata (Lamb.) Hook.) is an important fast-growing timber tree in southern China. Chinese fir has been grown for thousands of years in China and plays a significant role in both economic wood production and ecological functions, including carbon sequestration, soil and water conservation, climate regulation, and forest biodiversity. In the 21st century, the cultivation and efficient utilization of artificial timber stands has become the inevitable choice of China’s timber supply strategy. However, the average growth per unit area and stand quality of Chinese fir plantations are still very low. Research has explored the effects of planting density and site conditions on the timber assortment structure of Chinese fir plantations5,19, yet the effects of stand factors (e.g., the number of retained trees, stand density index, and the Gini coefficient) on the volume growth of trees of different diameter classes have not been studied under varying site conditions, planting density, and stand age. Furthermore, the effects of the interaction of site quality, planting density, and other stand factors on wood yield in trees of varying diameter classes remain to be explored. However, a thorough understanding of the impact of these factors on the timber assortment structure is useful for guiding the management of Chinese fir plantations and the oriented cultivation of target tree species, which may play an important role in realizing wood production potential and maintaining high wood yield. The purpose of this study was to determine the influence and relative importance of the initial planting density, site conditions, age, competition index (stand density index and number of living trees), and stand structural heterogeneity (the Gini coefficient) on total timber yield, and the large-, medium-, and small-diameter timber of Chinese fir plantations in the northwest of the Chinese fir distribution zone. More specifically, we examined the following several interrelated questions: (1) What effects do initial planting density and site quality have on the outturn of various timber assortments at different ages?; (2) What is the coupling relationship between various stand factors and timber assortment structure?; and (3) What is the relative importance of initial planting density, site conditions, age, competition, and stand structure heterogeneity on the outturn and outturn rate of different timber assortments? Materials and methods Study area and data collection This study was conducted in the Naxi District (28°47′ N, 105°23′ E), Luzhou City, Sichuan Province. The study site is located in the northwest of the Chinese fir distribution zone. The average annual temperature is 17.5 ℃ and the average annual precipitation is 1,182 mm. The topography of the study area is mainly high hills with an altitude of about 440 m, and the soil is primarily red soil developed on shale. A total of 15 experimental plots (20 m × 30 m) were established in the spring of 1982 with one-year-old bare-rooted Chinese fir seedlings in a completely randomized block design with three blocks. Each block consisted of five planting densities: A (1,667 trees/ha, 2 m × 3 m); B (3,333 trees/ha, 2 m × 1.5 m); C (5,000 trees/ha, 2 m × 1 m); D (6,667 trees/ha, 1 m × 1.5 m); and E (10,000 trees/ha, 1 m × 1 m). Each plot was surrounded by a 2-row buffer zone of similarly treated trees and all trees were tagged for continuous measurements. The diameter at breast height (DBH) was measured after the tree height reached 1.3 m. Stand dominant height was determined as the average height of the six tallest trees in each plot, with that at the reference age of 20 years taken as the site index. Trees were surveyed after they stopped growing at the end of the year or before they began to grow the following year. Measurements were performed every year from 1985 to 1995 and every two or three years from 1995 to 2013. The last survey was conducted in 2018. Calculation of the timber assortment outturn and outturn rate Based on the DBH size and the classification standard of each timber assortment, the number of timber assortment trees in each diameter class was calculated according to Table 129. Table 1Percentage of different kinds of lumber trees in each diameter class.Diameter class (cm)68101214161820222426Timber assortmentFWLNTLNTLNTSTLNTSTSTSTSTMTMTMTMTLTLTPercentage (%)30701006040109010010040601001006535100This table refers to the book “Study on Management Statistics and Tables and Optimal Density Control of Cunninghamia lanceolata Forest”29. The values in the table were calculated according to 1,094 felled trees. FW: fuelwood; LNT: long narrow timber (Wood less than 5 m in length or less than 7 cm in diameter inside bark at 2.5 m from base); ST: small-diameter timber (Tree-length with a diameter inside bark 7 ≤ D < 15 cm at 2.5 m from base); MT: medium-diameter timber (Tree-length with a diameter inside bark 15 ≤ D < 21 cm at 2.5 m from base); LT: large-diameter timber (Tree-length with a diameter inside bark D ≥ 21 cm at 2.5 m from base). This paper only uses dimension lumber (i.e., large-, medium-, and small-diameter timber) data.Formulas (1)–(3) were used to calculate the outturn for each diameter class of the dimension lumber (large-, medium- and small-diameter timber) of living wood: 1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\text{V}}_{{\text{t}}} = \:{\text{3}}.{\text{60243758}}\: \times \:{\text{10}}^{{ - {\text{5}}}} \: \times \:{\text{D}}^{{{\text{1}}.{\text{94752076}}\:}} \: \times \:{\text{H}}^{{{\text{1}}{\text{.00793769}}}} ,$$\end{document} 2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\text{V}}_{{\text{f}}} = {\text{5}}.{\text{35633889}}\: \times \:{\text{10}}^{{ - {\text{3}}}} \: \times \:{\text{D}}^{{ - {\text{0}}{\text{.567409453}}}} \: \times \:{\text{H}}^{{{\text{0}}{\text{.632966036}}}},$$\end{document} 3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\text{V}}_{{\text{w}}} = \:{\text{2}}.{\text{85679844}}\: \times \:{\text{10}}^{{ - {\text{5}}}} \: \times {\text{D}}^{{{\text{1}}{\text{.85964898}}}} \: \times \:{\text{H}}^{{{\text{0}}{\text{.686898895}}}},$$\end{document} Table 1Percentage of different kinds of lumber trees in each diameter class.Diameter class (cm)68101214161820222426Timber assortmentFWLNTLNTLNTSTLNTSTSTSTSTMTMTMTMTLTLTPercentage (%)30701006040109010010040601001006535100This table refers to the book “Study on Management Statistics and Tables and Optimal Density Control of Cunninghamia lanceolata Forest”29. The values in the table were calculated according to 1,094 felled trees. FW: fuelwood; LNT: long narrow timber (Wood less than 5 m in length or less than 7 cm in diameter inside bark at 2.5 m from base); ST: small-diameter timber (Tree-length with a diameter inside bark 7 ≤ D < 15 cm at 2.5 m from base); MT: medium-diameter timber (Tree-length with a diameter inside bark 15 ≤ D < 21 cm at 2.5 m from base); LT: large-diameter timber (Tree-length with a diameter inside bark D ≥ 21 cm at 2.5 m from base). This paper only uses dimension lumber (i.e., large-, medium-, and small-diameter timber) data.Formulas (1)–(3) were used to calculate the outturn for each diameter class of the dimension lumber (large-, medium- and small-diameter timber) of living wood: Percentage of different kinds of lumber trees in each diameter class. This table refers to the book “Study on Management Statistics and Tables and Optimal Density Control of Cunninghamia lanceolata Forest”29. The values in the table were calculated according to 1,094 felled trees. FW: fuelwood; LNT: long narrow timber (Wood less than 5 m in length or less than 7 cm in diameter inside bark at 2.5 m from base); ST: small-diameter timber (Tree-length with a diameter inside bark 7 ≤ D < 15 cm at 2.5 m from base); MT: medium-diameter timber (Tree-length with a diameter inside bark 15 ≤ D < 21 cm at 2.5 m from base); LT: large-diameter timber (Tree-length with a diameter inside bark D ≥ 21 cm at 2.5 m from base). This paper only uses dimension lumber (i.e., large-, medium-, and small-diameter timber) data. Formulas (1)–(3) were used to calculate the outturn for each diameter class of the dimension lumber (large-, medium- and small-diameter timber) of living wood: where Vt is the timber outturn (m3/ha); Vf is the fuel wood outturn (m3/ha); Vw is the waste wood outturn (m3/ha); D is the average DBH of each diameter class (cm); and H is the average height of each diameter class (m). Formulas (4)–(6) were used to calculate the long narrow timber outturn of living wood:4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\:{\text{V}}_{{{\text{t(LNT)}}}} = \: - {\text{0}}.{\text{0275552409}}\:{\text{ + }}\:{\text{3}}.{\text{68649463}}\: \times {\text{10}}^{{ - {\text{3}}}} \: \times \:{\text{D}}\:{\text{ + }}\:{\text{1}}.{\text{67244305}}\: \times {\text{10}}^{{ - {\text{3}}}} \: \times \:{\text{H}},$$\end{document}5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\text{V}}_{{{\text{f(LNT)}}}} = \:{\text{9}}.{\text{44348782}}\: \times {\text{10}}^{{-{\text{4}}}} \: - \:{\text{8}}.{\text{25571796}}\: \times \:{\text{10}}^{{ - {\text{5}}}} \: \times \:{\text{D}}\:{\text{ + }}\:{\text{3}}.{\text{39741113}}\: \times {\text{10}}^{{-{\text{4}}}} \: \times \:{\text{H}},$$\end{document}6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\text{V}}_{{{\text{w(LNT)}}}} = - {\text{9}}.{\text{71138159}}\: \times \:{\text{10}}^{{ - {\text{3}}}} \:{\text{ + }}\:{\text{1}}.{\text{67009235}} \times {\text{10}}^{{ - {\text{3}}}} \: \times \:{\text{D}}\:{\text{ + }}\:{\text{3}}.{\text{11969525}}\: \times {\text{10}}^{{-{\text{4}}}} \: \times \:{\text{H}},$$\end{document} where Vt(LNT) is the timber outturn (m3/ha); Vf(LNT) is the fuel wood outturn (m3/ha); Vw(LNT) is the waste wood outturn (m3/ha); D is the average DBH of each diameter class (cm); and H is the average height of each diameter class (m). The empirical formula for binary standing timber volume from the departmental standard of Chinese fir was used to calculate the tree volume:7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\text{V}} = \:{\text{0}}.{\text{000058777042}}\: \times \:{\text{D}}^{\text{1.9699831}}\: \times \:{\text{H}}^{\text{0.89646157}},$$\end{document} Formula (7) was employed to calculate the average volume per tree of each diameter class, which is then multiplied by the number of trees of each diameter class to obtain the total volume of each diameter class. The sum of the volume of all diameter classes, namely, the stand volume of each plot, is then multiplied by the conversion factor to obtain the stand volume per hectare. Based on the timber assortment type, the timber outturn of all diameter classes of the same timber assortment was merged to obtain the outturn of each timber assortment of the stand, including the timber outturn of large-, medium-, and small-diameter timber and long narrow timber, and the volume of fuel wood and waste wood. The outturn rate of each kind of timber assortment is calculated as the ratio of its outturn to the total outturn of all timber assortments. Stand factors To explore the influence of the stand factors on the timber assortment structure of Chinese fir plantations, we selected six variables: stand age (age), planting density (PD), site index (SI), stand density (SD), stand density index (SDI), and the Gini coefficient (Gini). PD was used to account for the effect of initial spacing on the timber assortment structure. SI reflects the site condition, and the larger the site class, the better the site quality. SD represents the stand density of stands in different periods and can be expressed as the number of living trees per hectare. SD is used to quantify the effect of competition intensity on productivity22. SDI is a comprehensive index of the average DBH and the number of trees. It can reflect both the number of trees and the tree size in a stand30, simultaneously describing the stand density and characterizing the stand structure. Thus, SDI can accurately describe stand density and is sensitive to the diameter class distribution31. SDI is calculated as follows: \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\text{SDI}=\text{N}\times{\left({\text{D}}_{\text{g}}/{\text{D}}_{\text{0}}\right)\:}^{\beta}$$\end{document}, where N is the number of trees per unit area; Dg is the square mean diameter; D0 is the standard mean diameter (Chinese fir is 20 cm); and β is the self-thinning rate of the stand. The N-Dg relation (logN=a + βlogDg) can be used to calculate the stand’s self-thinning rate (β) when it reaches the full stocking. The Gini coefficient is used to quantify stand structural diversity26,27 and was selected here to measure the stand structural heterogeneity. It can reflect the individual size differentiation of trees in the stand. The Gini coefficient ranges from 0 to 1, and the closer it is to 0, the smaller the size difference among trees. Conversely, values closer to 1 indicate a greater size diversity. The Gini coefficient was calculated using the package “ineq”32in R33 to represent the degree of diameter inequality in a stand. Statistical analysis A two-way ANOVA was used to examine the effects of planting density, site index, and their interactions on different timber assortment outturns using SPSS Statistics 25.0. Structural equation modeling (SEM) was employed to identify the relationship between the outturn of different timber assortments and stand age (age), initial planting density (PD), site index (SI), stand density (SD), the Gini coefficient (Gini), and stand density index (SDI) using Amos 24.0. To quantify the relative importance of these six stand factors to different timber assortment outturns and outturn rates of dimension lumber, hierarchical partitioning (HP) analysis was used to calculate the relative contributions of the six stand factors with the hier.part package in R34. HP analysis can be used to rank the importance of variables under the influence of multiple factors. It determines the influence degree of each factor on the dependent variable through the hierarchical decomposition of the influencing factors of the dependent variable and subsequently realizes the relative importance ranking of the factors35. Results Effects of site quality on the timber assortment outturn and outturn rate As the stand age progressed, the total timber yield of all density plots exhibited a rapid early increase followed by a steady increase and a minor decline between the ages of 24 and 33, and subsequently increased again. Prior to the age of 24 years, the growth trajectory of total timber yield was essentially the same as that of the small-diameter timber. The higher the site quality, the faster the growth rate in the early stages and larger the growth increment. The emergence time of large-diameter timber decreased with the increase in site class. The medium-diameter timber outturn and outturn rate exhibited an upward trend with the stand age and site class. During whole the study period, the medium-diameter timber outturn increased continuously, while the large-diameter timber was consistently low. The outturn and outturn rates of medium-diameter timber under the same planting density were greater than those of large-diameter timber in all site classes. The outturn and outturn rate of small-diameter timber tended to initially increase and then decrease with stand age. Large variations were observed between the site classes of 12 and 14 in 15 to 25 years (Fig. 1). Fig. 1Dynamic changes of different timber assortment outturn and outturn rate under the different site classes: (a) timber assortment outturn in B density (3,333 trees/ha); (b) timber assortment outturn in D density (6,667 trees/ha); (c) timber assortment outturn rate in B density; (d) timber assortment outturn rate in D density. Total, total timber outturn or outturn rate of dimension lumber (sum of the outturn or outturn rate of large-, medium-, and small-diameter timber); Large, large-diameter timber; Medium, medium-diameter timber; Small, small-diameter timber; SI, site index. Dynamic changes of different timber assortment outturn and outturn rate under the different site classes: (a) timber assortment outturn in B density (3,333 trees/ha); (b) timber assortment outturn in D density (6,667 trees/ha); (c) timber assortment outturn rate in B density; (d) timber assortment outturn rate in D density. Total, total timber outturn or outturn rate of dimension lumber (sum of the outturn or outturn rate of large-, medium-, and small-diameter timber); Large, large-diameter timber; Medium, medium-diameter timber; Small, small-diameter timber; SI, site index. Effects of initial planting density on timber assortment outturn and outturn rate For site index class 12 (Fig. 2a and c), as the planting density rose from B to E in 7 to 22 years, the total timber yield and small-diameter timber outturn decreased. After 22 years, the high densities gradually exceeded the low densities, and the total timber yield and small-diameter timber outturn increased with planting density. The total timber yield rate of the dimension lumber increased with the decrease in planting density, and the lower the planting density, the earlier the total timber yield rate approached the stable maximum. As age increased, the total timber yield rate of all densities eventually approached 65–70%. The density effect of the small-diameter timber outturn rate was close to that of the outturn, yet the positive density effect of the outturn rate did not appear until the age of 28 years. The medium-diameter timber outturn and outturn rate exhibited a negative correlation with planting density. In all densities of site class 12, large-diameter timber appeared only at the last observation, and there were no obvious differences between densities. For site index class 14 (Fig. 2b and d), the total timber yield increased with planting density after 24 years. A lower planting density indicated a higher total timber yield rate for dimension lumber. The total timber yield rate of all density plots was close to 70–80%. The peak value of the small-diameter timber outturn increased with planting density. The peak value of the small-diameter timber outturn with densities B and D was greater than that of density A, yet there were no evident differences between densities B and D. Conversely, the lower the planting density, the higher the peak value of the outturn rate of the small-diameter timber, and the shorter the time taken to reach its peak value. The medium-diameter timber outturn and outturn rate decreased as planting density increased. Moreover, the outturn and outturn rate of medium-diameter timber with density A were markedly higher than those of densities B and D during the ages from 15 to 30 years. Large-diameter timber appeared in densities A, B, and D at the ages 19, 24, and 38, respectively. At 38 years, the outturn and outturn rate of large-diameter timber decreased as planting density increased. Fig. 2Dynamic changes of different timber assortment outturn and outturn rate under the different initial planting densities: (a) timber assortment outturn in site index class 12; (b) timber assortment outturn in site index class 14; (c) timber assortment outturn rate in site index class 12; (d) timber assortment outturn rate in site index class 14. A, 1,667 trees/ha; B, 3,333 trees/ha; C, 5,000 trees/ha; D, 6,667 trees/ha; E, 10,000 trees/ha. The mean yield of each kind of timber assortment in plots B1 and B2 from the same site index class is shown by B1 + 2. This also applies to other symbols. Total, total timber yield or yield rate; Large, large-diameter timber; Medium, medium-diameter timber; Small, small-diameter timber; SI, site index. Dynamic changes of different timber assortment outturn and outturn rate under the different initial planting densities: (a) timber assortment outturn in site index class 12; (b) timber assortment outturn in site index class 14; (c) timber assortment outturn rate in site index class 12; (d) timber assortment outturn rate in site index class 14. A, 1,667 trees/ha; B, 3,333 trees/ha; C, 5,000 trees/ha; D, 6,667 trees/ha; E, 10,000 trees/ha. The mean yield of each kind of timber assortment in plots B1 and B2 from the same site index class is shown by B1 + 2. This also applies to other symbols. Total, total timber yield or yield rate; Large, large-diameter timber; Medium, medium-diameter timber; Small, small-diameter timber; SI, site index. Relationships between the planting density, site quality, and critical time nodes for the timber assortment At the same initial planting density, the larger the site class, the higher the maximum value of total timber yield, medium- and small-diameter timber outturn, mean annual increment (MAI), and current annual increment (CAI). The maximum value and age at which the total timber yield and small-diameter timber outturn were maximized were positively correlated with initial planting density. For example, at site class 14, the small-diameter timber outturn of densities A, B, and D peaked at 94.58 m3/ha, 137.17 m3/ha, and 141.97 m3/ha at ages 14, 22, and 24, respectively. At the same site class, the total timber yield and small-diameter timber outturn of density B exhibited the highest maximum MAI and CAI values. In addition, the emergence age of medium-diameter timber was positively correlated with the initial planting density (Table 2). Table 2Primary time nodes for the growth of total timber yield, and medium-, and small-diameter timber in different planting densities and site classes.Timber assortment typePlanting density (trees/ha)Site class (m)Emergence age (year)Maximum timber outturn (m3/ha)Maximum timber outturn age (year)Maximum MAI for timber outturn (m3/ha)Maximum CAI for timber outturn (m3/ha)Total timber yield1,667147144.453818.7132.813,333127116.103815.4326.363,333147170.643822.7441.755,000127126.893812.0220.206,667127129.52387.9210.176,667147196.163815.2324.9910,000128146.44387.609.41Medium-diameter timber1,66714966.86383.715.673,333121044.22381.523.283,333141350.56381.813.745,000121339.58381.413.216,667121923.59381.523.096,667141567.47382.815.8510,000121937.02381.762.77Small-diameter timber1,66714794.581418.1032.813,33312792.212615.2126.363,333147137.172222.7441.755,000127106.422412.0220.206,667127113.18267.9210.176,667147141.972415.2324.9910,000128105.01287.609.41 Table 2Primary time nodes for the growth of total timber yield, and medium-, and small-diameter timber in different planting densities and site classes.Timber assortment typePlanting density (trees/ha)Site class (m)Emergence age (year)Maximum timber outturn (m3/ha)Maximum timber outturn age (year)Maximum MAI for timber outturn (m3/ha)Maximum CAI for timber outturn (m3/ha)Total timber yield1,667147144.453818.7132.813,333127116.103815.4326.363,333147170.643822.7441.755,000127126.893812.0220.206,667127129.52387.9210.176,667147196.163815.2324.9910,000128146.44387.609.41Medium-diameter timber1,66714966.86383.715.673,333121044.22381.523.283,333141350.56381.813.745,000121339.58381.413.216,667121923.59381.523.096,667141567.47382.815.8510,000121937.02381.762.77Small-diameter timber1,66714794.581418.1032.813,33312792.212615.2126.363,333147137.172222.7441.755,000127106.422412.0220.206,667127113.18267.9210.176,667147141.972415.2324.9910,000128105.01287.609.41 Primary time nodes for the growth of total timber yield, and medium-, and small-diameter timber in different planting densities and site classes. Effects of the site index and initial planting density and their interaction on timber assortment outturn under different stand ages The two-way ANOVA results (Table 3) showed that the density and site effects significantly affected the total timber yield and small-diameter timber outturn between 8 and 10 years and 8 to 24 years (P < 0.05), respectively. At the last observation (38 years old), the outturn of large-diameter timber differed significantly depending on planting density and the site index (P < 0.05). Prior to this age, there were no statistically significant differences between the density effect and site effect of all stand ages. Planting density was found to have a significant influence on medium-diameter timber outturn from 9 to 30 years (P < 0.05). During the whole observation period, the interaction between planting density and site index had no significant effect on the outturn of all timber assortments. Table 3Significance test of initial planting density and the site index on different timber assortment outturns under different ages.Stand ageTotal timber yieldLarge-diameter timberMedium-diameter timberSmall-diameter timberPlanting densitySite indexDensity × sitePlanting densitySite indexDensity × sitePlanting densitySite indexDensity × sitePlanting densitySite indexDensity × site73.1470.0510.2823.1470.0510.28286.434*8.925*0.0246.434*8.925*0.02495.485*7.959*0.15993.203***0.0000.0005.498*8.005*0.160104.040*7.563*0.0281.4000.0700.0703.962*7.722*0.016131.6298.117*0.0302.2300.0820.0822.0609.560*0.011141.2427.435*0.0013.7170.0050.0051.8359.425*0.000151.2828.684*0.0066.043*0.4720.0222.20310.123*0.003171.2438.917*0.00912.825**0.7390.0452.2949.882*0.004190.9308.313*0.0410.4000.0000.00049.193***1.9350.0932.5519.349*0.034220.6317.505*0.0401.0580.0000.00031.114***3.1310.0003.8148.339*0.057240.4475.974*0.1170.4630.4090.40912.301**0.8940.0193.3566.010*0.046260.5184.6490.0020.4700.4410.4417.046*0.9211.9313.3224.0950.081280.5384.1720.0020.4680.4850.4857.194**1.2302.2993.5413.3080.095300.5114.5420.0210.7432.0542.0544.613*1.7202.7003.3022.7760.240330.7535.244*0.0601.4501.1001.1001.3212.9262.4383.3941.9030.502381.3878.079*0.0804.976*7.688*1.5910.1894.7272.6425.854*2.5950.251The table presents the F-ratio values. Asterisk indicates statistically significant differences *P < 0.05, **P < 0.01, ***P < 0.001. Table 3Significance test of initial planting density and the site index on different timber assortment outturns under different ages.Stand ageTotal timber yieldLarge-diameter timberMedium-diameter timberSmall-diameter timberPlanting densitySite indexDensity × sitePlanting densitySite indexDensity × sitePlanting densitySite indexDensity × sitePlanting densitySite indexDensity × site73.1470.0510.2823.1470.0510.28286.434*8.925*0.0246.434*8.925*0.02495.485*7.959*0.15993.203***0.0000.0005.498*8.005*0.160104.040*7.563*0.0281.4000.0700.0703.962*7.722*0.016131.6298.117*0.0302.2300.0820.0822.0609.560*0.011141.2427.435*0.0013.7170.0050.0051.8359.425*0.000151.2828.684*0.0066.043*0.4720.0222.20310.123*0.003171.2438.917*0.00912.825**0.7390.0452.2949.882*0.004190.9308.313*0.0410.4000.0000.00049.193***1.9350.0932.5519.349*0.034220.6317.505*0.0401.0580.0000.00031.114***3.1310.0003.8148.339*0.057240.4475.974*0.1170.4630.4090.40912.301**0.8940.0193.3566.010*0.046260.5184.6490.0020.4700.4410.4417.046*0.9211.9313.3224.0950.081280.5384.1720.0020.4680.4850.4857.194**1.2302.2993.5413.3080.095300.5114.5420.0210.7432.0542.0544.613*1.7202.7003.3022.7760.240330.7535.244*0.0601.4501.1001.1001.3212.9262.4383.3941.9030.502381.3878.079*0.0804.976*7.688*1.5910.1894.7272.6425.854*2.5950.251The table presents the F-ratio values. Asterisk indicates statistically significant differences *P < 0.05, **P < 0.01, ***P < 0.001. Significance test of initial planting density and the site index on different timber assortment outturns under different ages. The table presents the F-ratio values. Asterisk indicates statistically significant differences *P < 0.05, **P < 0.01, ***P < 0.001. Coupling relationship between the timber assortment outturn and stand factors Stand factors are important in regulating the timber assortment structure and limiting timber assortment growth. The direct, indirect, and total effects of six stand factors on the outturn of different timber assortments were analyzed using the structural equation model (Fig. 3). The direct effects of age and SI on total timber yield (TODL) were 0.38 and 0.17, respectively. Both of them had significant indirect effects on TODL, with indirect effects of 0.41 and 0.08, respectively. The direct effect of PD on TODL was − 0.06, with a relatively low correlation between them. In contrast, PD exerted a significant indirect effect on TODL by influencing SDI, SD, and Gini, with an indirect effect of − 0.12 (Fig. 3a). Age, PD, SI, SDI, and SD all had a significant direct impact on large-diameter timber outturn (LTO), with direct effect coefficients of 0.74, − 0.50, 0.28, − 0.18, and 0.66, respectively. Moreover, age, PD, and SI exerted significant indirect effects on LTO, with indirect effects of − 0.10, 0.31, and − 0.06. However, their influence on the growth of large-diameter timber was primarily due to their direct effects (Fig. 3b). Age, SI, and SD had a positive effect on medium-diameter timber outturn (MTO), while PD and SDI had a negative effect. Age, PD, and SI had significant indirect effects on the MTO, with indirect effect coefficients of − 0.08, 0.17, and − 0.06, respectively. Furthermore, age, PD, and SI primarily influenced the growth of medium-diameter timber through their direct effects (Fig. 3c). SDI, age, and SD had the greatest total effects on small-diameter timber outturn (STO), with influence coefficients of 0.77, 0.64, and − 0.59, respectively. Age, PD, and SI had significant indirect effects on the STO by influencing SDI, SD, and Gini. Furthermore, the influence of age, PD, and SI on the growth of small-diameter timber was primarily due to indirect effects via other factors (Fig. 3d). The findings revealed that adjusting the number of retained trees and the degree of individual size differentiation could enhance the effects of age, PD, and SI on small-diameter timber. Fig. 3Path diagrams for the influence of stand factors on the (a) total timber yield of dimension lumber (TODL), (b) large-diameter timber outturn (LTO), (c) medium-diameter timber outturn (MTO), and (d) small-diameter timber outturn (STO). Lines represent the effect of the exogenous variables on endogenous variables, while the numbers on the lines are the standardized direct path coefficients. The numbers in red are the standardized indirect path coefficients of age, PD, and SI for different timber assortment outturns. SDI, stand density index; SD, stand density; Gini, Gini coefficient; Age, stand age; PD, planting density; SI, site index. *P < 0.05, **P < 0.01, ***P < 0.001. Path diagrams for the influence of stand factors on the (a) total timber yield of dimension lumber (TODL), (b) large-diameter timber outturn (LTO), (c) medium-diameter timber outturn (MTO), and (d) small-diameter timber outturn (STO). Lines represent the effect of the exogenous variables on endogenous variables, while the numbers on the lines are the standardized direct path coefficients. The numbers in red are the standardized indirect path coefficients of age, PD, and SI for different timber assortment outturns. SDI, stand density index; SD, stand density; Gini, Gini coefficient; Age, stand age; PD, planting density; SI, site index. *P < 0.05, **P < 0.01, ***P < 0.001. Relative importance of various stand factors in determining the growth of timber assortment Variation of the relative contributions of stand factors on timber assortment structure The relative contributions of Gini, PD, and SD to TODL were observed to decrease each year as the stand age increased. However, the relative contribution of SDI to TODL increased with age, surpassing those of other stand factors around the age of 20. The relative contribution of SI to TODL exhibited minimal fluctuations with age, ranging from 20.35 to 34.75%. The relative contribution of SI to LTO and LTOR decreased with age and subsequently increased slightly after 33 years. Before 28 years, SI exhibited the highest relative contribution rate to LTO and LTOR. After the age of 28, Gini had a higher relative contribution rate than SI. Throughout the observation period, the relative contribution rate of PD, SD, and SDI to LTO and LTOR was consistently low, with rates below 22%. SI and SDI contributed the most to MTO and MTOR in the early stages of medium-diameter timber formation, and their contributions then decreased each year as the stand age increased. The contribution rate of SI and SDI to MTO reached its lowest point around 22–24 years old and then gradually increased. The relative contributions of PD and SD to MTO and MTOR initially increased and subsequently decreased with age, with SD demonstrating the highest relative contribution rate to MTO and MTOR after 17 years. SDI had the lowest relative contribution rate of all the stand factors in the first 7–10 years after the formation of small-diameter timber. Following this, the relative contribution rate of SDI gradually increased, peaking at the age of 22 at 70.70%. SDI had the highest relative contribution rate of any stand factor between 13 and 30 years. After 30 years, the relative contribution rate of SD surpassed that of SDI, making it the most influential variable for STO. Moreover, the relative contribution rate of SD to STOR initially increased and subsequently decreased with age until it reached its lowest point (17.95% at 24 years old) and then steadily increased. After 30 years, it was the highest among all stand factors (Fig. 4). Fig. 4Dynamic changes in the relative contributions of stand factors to the timber assortment outturn and outturn rate. TODL total timber yield of dimension lumber, TORDL total timber outturn rate of dimension lumber, LTO large-diameter timber outturn, LTOR large-diameter timber outturn rate, MTO medium-diameter timber outturn, MTOR medium-diameter timber outturn rate, STO small-diameter timber outturn, STOR small-diameter timber outturn rate, SDI stand density index, SD stand density, Gini Gini coefficient, Age stand age, PD planting density, SI site index. Dynamic changes in the relative contributions of stand factors to the timber assortment outturn and outturn rate. TODL total timber yield of dimension lumber, TORDL total timber outturn rate of dimension lumber, LTO large-diameter timber outturn, LTOR large-diameter timber outturn rate, MTO medium-diameter timber outturn, MTOR medium-diameter timber outturn rate, STO small-diameter timber outturn, STOR small-diameter timber outturn rate, SDI stand density index, SD stand density, Gini Gini coefficient, Age stand age, PD planting density, SI site index. Relative importance of stand factors on timber assortment structure SDI had the greatest impact on total timber yield (TODL) and small-diameter timber (STO and STOR), followed by age and SD. However, the contribution rates of PD, SI, and Gini to total timber yield and small-diameter timber were relatively small, accounting for less than 10%. Age was the most important factor for large- (LTO and LTOR) and medium- (MTO and MTOR) diameter timber, followed by SI and SD for LTO and LTOR, SD and SI for MTO, and SD and PD for MTOR, respectively. In contrast, SDI and Gini made relatively small contributions to large- and medium-diameter timber, with relative contribution rates of less than 6% (Fig. 5). Fig. 5Relative contributions of stand factors to the timber assortment outturn and outturn rate. TODL total timber yield of dimension lumber, TORDL total timber outturn rate of dimension lumber, LTO large-diameter timber outturn, LTOR large-diameter timber outturn rate, MTO medium-diameter timber outturn, MTOR medium-diameter timber outturn rate, STO small-diameter timber outturn, STOR small-diameter timber outturn rate, SDI stand density index, SD stand density, Gini Gini coefficient, Age stand age, PD planting density, SI site index. Relative contributions of stand factors to the timber assortment outturn and outturn rate. TODL total timber yield of dimension lumber, TORDL total timber outturn rate of dimension lumber, LTO large-diameter timber outturn, LTOR large-diameter timber outturn rate, MTO medium-diameter timber outturn, MTOR medium-diameter timber outturn rate, STO small-diameter timber outturn, STOR small-diameter timber outturn rate, SDI stand density index, SD stand density, Gini Gini coefficient, Age stand age, PD planting density, SI site index. Discussion The total timber yield of dimension lumber and the outturn of large-, medium-, and small-diameter timber increased with the site class (Fig. 1). This is consistent with previous research that reports an improvement in wood production due to enhanced site quality15,36. Large-diameter timber trees appeared earlier in sites with higher site quality and lower planting densities. At site class 12, large-diameter timber began to appear around the age of 38, while at site class 14, density plots A, B, and D began to appear at 19, 24, and 38 years (Fig. 2). In a previous study, we explored the role of site quality and planting density on the timber assortment structure of Chinese fir in the east of the Chinese fir distribution zone. The results revealed the significant effect of site on the formation of large-diameter timber, and better site quality shortened the formation time of large-diameter timber. The optimal site conditions for cultivating Chinese fir large-diameter timber trees were observed to include site classes ≥ 20, a planting density of no more than 1,667 trees/ha, and a rotation age of more than 26 years37. However, each of the experimental forest plots of site classes 12 and 14 produced minimal large-diameter timber. The findings indicated that the current site conditions in this climate zone were unsuitable for growing large-diameter timber trees. Hui et al. (2000)38 found that the proportion of medium- and large-diameter timber in the plots below site class 14 was extremely low, regardless of density. Large-diameter timber yield was also low at site class 16, yet a comparatively large proportion of medium-diameter timber could be produced. Hui et al. suggested that sites with site class 16 or more are appropriate for cultivating medium-diameter timber. The site conditions of the current study (site classes 12 and 14) are suitable for growing small-diameter timber trees. At site class 14, the initial planting density was less than 1,667 trees/ha, indicating that medium-diameter timber could be grown, but the cultivation period should be extended. In particular, at the age of 33, the yield curve of the small-diameter timber of 1,667 trees/ha intersected the yield curve of the medium-diameter timber, and after 33 years, the yield of the medium-diameter timber continued to rise (Fig. 2b and d). Thus, a significant proportion of the medium-diameter timber could be obtained if the cultivation was long enough. At a low initial planting density, medium- and large-diameter timber yielded more and grew faster, while small-diameter timber yield peaked earlier. This is consistent with the findings of Liu et al. (2019)13. The initial planting density determines the intensity of resource competition among individual trees. At higher planting densities, individual trees grow slowly or die due to intense intraspecific competition at high plantation densities. Low planting densities may allow trees to receive more light, water, and nutrients, which will increase DBH, height, and volume10. The density effect on the total timber yield of dimension lumber and the outturn of small-diameter timber demonstrated both a negative inhibition and positive promotion during various growth stages. The total timber yield and small-diameter timber outturn were negatively impacted by density before the age of 22. Furthermore, the density exerted a positive impact on the total timber yield and small-diameter timber outturn after 22 years. The density-yield relationship at different growth stages may be related to self-thinning of stands. Self-thinning of stands can cause changes in stand density, altering the trajectory of the density-wood production relationship. Yang et al. (2019)39 demonstrated that the effects of stand density on productivity varied with the growth stage due to both positive promotion and negative inhibition. As the effect of density on wood production varies with the growth and development stages, different growth and development stages should be considered when analyzing the effect of density on wood production. Forest productivity is closely related to stand structure. Previous studies have shown that stand age, stand density, and diversity have a great impact on forest productivity23,24. At the same time, a single stand structure attribute may indirectly affect forest productivity by influencing other stand structure attributes. More specifically, forest productivity is the result of the coupling effect of many factors23. SD is a competition index that can describe stand density and quantify the impact of competition intensity on stand growth or productivity22. SDI is a good measure of stand crowding and can effectively describe competition as it integrates tree size and the number of trees in stands into a single parameter30. In this study, SDI, age, and SD were the most important factors influencing both total timber yield and small-diameter timber, contributing 89.37% to total timber yield and 92.37% to small-diameter timber outturn. The negative effect of SD on total timber yield and small-diameter timber, as well as the positive effect of SDI on both, suggest that wood yield is determined by the combined effect of tree number and tree size distribution, and that size-related increases in growth (growth increment and growth rate) will compensate for the increase in mortality caused by density-related competition. Harrington et al. (2009)8 reported that SDI should be kept within a certain threshold as competition-related deaths increase once that threshold is exceeded. Therefore, further research on the relationship between tree size distribution and asymmetric competition is required to explain the mechanism of wood yield change in terms of growth and mortality40. The effects of age, PD, and SI on total timber yield and small-diameter timber were found to be enhanced by adjusting the number of retained trees and the differentiation degree of individual trees. Yan et al. (2023)41 found that stand age and planting density had an indirect effect on stand productivity, primarily by influencing the number of living trees or the Gini coefficient. The authors also revealed that planting density had the greatest effect on individual size differentiation (Gini), and choosing an appropriate planting density to reduce individual differentiation is critical for regulating stand productivity. Sun et al. (2018)27 concluded that initial planting density and site quality were the primary factors influencing pure stand size inequality, productivity, and natural sparseness. Density-dependent mortality can reduce size inequality and alter the relationship between size inequality and productivity as age increases. These results highlight the importance of considering the effects of the planting density and site index on SDI, SD, and Gini when analyzing the relationship between the planting density and site index and the total timber yield or small-diameter timber. Age was observed to exert the greatest relative contribution to large-diameter timber, followed by SI and SD (Fig. 5). Moreover, SI contributed the most to large-diameter timber until the age of 28, after which Gini was then identified as the main influencing factor (Fig. 4). Around the age of 28, Chinese fir plantations enter the continuous natural thinning stage, which is the continuation of the natural thinning peak period. At this time, there is still a high level of intraspecific competition42. This increases the degree of size differentiation among individual trees and can lower the large-diameter timber outturn. This demonstrates the importance of site quality in the formation and growth of large-diameter timber in the early stage, while stand structure becomes more important in large-diameter timber growth as the stand structure changes in the later stages. When the individual size differentiation of a stand reaches a certain point, the stand structure must be modified through more human intervention (via thinning and tending) to decrease the amount of dead wood and accelerate the stand’s passage through the natural thinning stage. Moreover, cutting down pressed wood can reduce individual differentiation to maintain stand productivity while also serving as an important nutrient source for dominant tree growth and soil fertility41. The most influential factors for the medium-diameter timber outturn were age, SD, and SI, with relative contributions of 50.42%, 19.34%, and 11.60% (Fig. 5). These findings highlight the importance of site quality, rotation control, and density control in the cultivation of large- and medium-diameter timber. The timber yield of dimension lumber in the northwest of the Chinese fir distribution zone is primarily composed of small- and medium-diameter timber according to the current growth situation of each timber assortment type in the region. However, the yield from large-diameter timber is low. Thus, small- and medium-diameter timber should be the primary target timber assortments in the northwest of the Chinese fir distribution zone. Numerous studies have demonstrated that a plantation should be thick if its goal is to cultivate small-diameter wood, and sparse if its goal is to cultivate medium- or large-diameter wood7,13. Further research has explored these topics for actual forest management to determine how much to plant per unit area, when to harvest, how much should be harvested under varying site conditions, and so on5,19. Our findings indicate small-diameter timber as the target timber assortment for cultivation in site class 12. When the initial planting density was greater than or equal to 3,333 trees/ha, the rotation age ranges between 22 and 28 years. Higher planting density leads to a longer rotation age and a harvest yield of 100–120 m3/ha during main cutting. Plots with site class 14 are suitable for small-diameter timbers. Medium-diameter timbers can also be cultivated with an appropriate density and a long enough cultivation period. If medium-diameter timber is the first cultivated timber assortment, the initial planting density should be low (less than or equal to 1,667 trees/ha). However, when small-diameter timber is the first cultivated timber assortment, the initial planting density can exceed 1,667 trees/ha, with a rotation age between 20 and 25 years and a yield of about 140 m3/ha harvested during the main cutting. If the initial planting density is less than 1,667 trees/ha, the rotation period can be appropriately shortened, yielding approximately 100 m3/ha of small-diameter timber at approximately 15 years of age. Site quality is critical in the formation of the timber assortment as it directly influences the formation of the timber assortment, the formation time, and the amount of wood produced. When site quality is no longer a barrier to producing the desired timber assortment, initial planting density can become a major limiting factor5. Studies have shown that the stand structure and differentiation characteristics in plantations have a significant impact on forest productivity20,21,24. Our findings revealed multiple stand factors to have a significant impact on the growth of the timber assortment of dimension lumber, both directly and indirectly. Age, PD, and SI primarily influenced the growth of large- and medium-diameter timber through direct effects, whereas their influence on the growth of small-diameter timber was primarily due to indirect effects via other factors. This highlights the importance of considering the interactions between stand age, planting density, site conditions, and other stand structure attributes when cultivating the target timber assortment of Chinese fir. Tending and thinning can effectively control the stand structure. Selective thinning can increase the supply of habitat resources while also promoting the growth of remaining trees. However, thinning intensity must be accounted for. In particular, the thinning intensity is high, the stand takes several years to recover to its pre-thinning state and may even stay below pre-thinning levels43. As a result, excessive reductions in stand density can lead to suboptimal total stand production. However, excessive density can also result in suboptimal yield. There is a direct relationship between stand density and yield, and quantifying this relationship is critical for understanding the growth and economic impact of thinning treatments9. These findings indicate that in order to cultivate and manage the target timber assortment, it is first necessary to determine the optimal site conditions and planting density, followed by the maximum density for each age stage and the corresponding yield relationship. The stand structure and individual differentiation traits should then be modified to prevent wood yield from declining, with density management techniques including tending and thinning then used to regulate the stand density and competition. This makes it possible to build a healthy population structure, guarantee the stand’s production potential, enhance the quality of the wood, and ensure the wood yield during the key cutting time. Conclusion We found that age, planting density, site conditions, and competition exerted varying degrees of influence on the growth trajectory and wood yield of different timber assortments in the northwest of the Chinese fir distribution zone. The competition indices (SDI and SD) and age were the most important factors in determining total timber yield and small-diameter timber, contributing 89.37% to total timber yield and 92.37% to small-diameter timber outturn. Site quality was key in the early stages of large-diameter timber formation and growth, and stand structural heterogeneity became more important as the stand structure changes in the later stages. The effects of age, PD, and SI on small-diameter timber were primarily attributed to their indirect effects via SDI, SD, and Gini. Adjusting the number of living trees and the degree of individual size differentiation enhanced the effects of age, PD, and SI on small-diameter timber outturn and the outturn rate. The influence of age, PD, and SI on large- and medium-diameter timber was primarily due to their direct effects. The findings indicate that small- or medium-diameter timber should be produced in this area rather than large-diameter timber. In site class 12, the target timber assortment for production was small-diameter timber. When the planting density was greater than or equal to 3,333 trees/ha, the rotation age ranged between 22 and 28 years. A higher planting density resulted in a longer rotation age and a harvest yield of 100–120 m3/ha during the main cutting periods. Plots with site class 14 are suitable for small- and medium-diameter timbers. If medium-diameter timber is the first cultivated timber assortment, planting density should be limited to no more than 1,667 trees/ha. If small-diameter timber is the first cultivated timber assortment, the planting density can be more than 1,667 trees/ha, with a rotation age between 20 and 25 years, and a yield of about 140 m3/ha harvested during the main cutting period.