Effects of thinning on tree growth and soil physiochemical properties in Cunninghamia lanceolata plantation

ABSTRACT Tree growth, along with soil properties, is greatly affected by forest management. We used a typical sampling to study the impact of four thinning intensities (T1: 0%, 2500 stems ha−1; T2: 20%, 2010 stems ha−1; T3: 30%, 1750 stems ha−1; T4: 40%, 1500 stems ha−1) on the tree growth and soil physicochemical properties and their correlation in Cunninghamia lanceolata plantations. The average annual increments in tree height, diameter at breast height (DBH), and volume increased with thinning intensity, and those of T4 differed significantly (P < 0.05) from those of T1. The average annual stand volume increments of T4 were significantly (P < 0.05) lower than that of T1, while the maximum value presented at T3. However, the effect of thinning in promoting the growth of Chinese fir diminished with time. As the thinning intensity increased, the diameter class distribution of the sample stands moved rightwards. Moreover, thinning improved soil physiochemical properties. The effects of thinning on soil properties in 0–20 cm soil layer were greater than those in 20–40 cm soil layer. There was a positive correlation between available nitrogen, available potassium and tree growth. The results of this study showed that thinning had a potential effect on tree growth and soil properties. The heavy thinning intensity (approximately 1500 stems ha−1) was the optimum for maintaining economic and ecological benefits. However, heavy thinning significantly reduced stand volume. From the perspective of improving stand volume and biomass, a moderate thinning intensity (approximately 1750 stems ha−1) could be considered for adoption.


Introduction
Thinning, a major measure for improving forest quality and efficiency (Sheng 2001), can reduce the competition among trees, improve the canopy space and understory microenvironment, and regulate the synergy between plants and soil, thereby promoting tree growth and improving forestland quality Dang et al. 2018). For example, studies have shown that appropriate thinning can improve the growth conditions and health of forest trees and increase the sunlight-exposed canopy area and the diameter at breast height (DBH) and volume of individual trees (Zhang et al. 2005). Studies on the effect of thinning on stand volume have arrived at inconsistent conclusions. Previous studies suggested thinning led to significant increases in stand volume (Zhou et al. 2016;Jiang et al. 2019), while other studies suggested that thinning led to insignificant variations or even decreases in stand volume (Lindgren and Sullivan 2013;Li et al. 2014). For the effects of thinning on soil physiochemical properties, some studies have shown that thinning deteriorates soil properties, which leads to an increase in soil bulk density, a reduction in soil water holding capacity and porosity in a mixed natural forest in southeastern China (Zhou et al. 2015), and a decrease of nitrogen, phosphorus, and potassium in a mixed plantation in northeastern China ) and a Mediterranean forest in southeastern Spain (Baena et al. 2013). Other studies showed that thinning has little effect on soil physical and chemical properties at Coram Experimental Forest in northwestern Montana (Jang et al. 2016). The study of a Chinese pine plantation by Dang et al. (2018) showed that soil nutrient content increased with thinning intensity. The study of Larix principisrupprechtii by Zhang et al. (2019) found that a remaining density of 1400 stems ha −1 facilitated soil nutrient accumulation and improved soil air permeability. The study of Chinese fir plantation by Xu et al. (2020) showed that high intensity thinning (approximately 1600 stems ha −1 ) could facilitated the development of understory vegetation and improved forestland soil fertility. The inconsistencies between the above research results may be due to the different climatic conditions, vegetation types, soil types, and forest ages of the studies areas (Wolk and Rocca 2009;Suchara et al. 2021).
Chinese fir (Cunninghamia lanceolata) is a fast-growing species of timber tree widely cultivated in southern China. According to the Ninth National Forest Resources Inventory Report, the area of Chinese fir plantations in China has reached 9.90 × 10 6 ha (SFA. 2019). However, it has been found that the stand volume and timber production have not been satisfactory, even though the areas of the plantations continue to increase (Sheng 2018). This is due to the monocultures and high-density management practices, which are in place at the present time (Tian et al. 2011). Based on the important ecological and economic value of Chinese fir, many scholars have carried out extensive and indepth studies on Chinese fir plantation, including the effects of thinning and continuous planting on Chinese fir plantation, analysis of soil nutrient status, biomass and degradation cause of Chinese fir plantation under different planting density or different age (Sheng et al. 2003;Sun et al. 2015;Chao et al. 2015;Ma et al. 2015;Chen et al. 2018). Typically, Chinese fir plantations used for industrial timber production are thinned once or twice during a rotation period (approximately 25 years) (Deng et al. 2009;Xu et al. 2014).
Traditionally, various studies of thinning have provided inconsistent results on soil properties or understory because focused on different tree species and conducted on different area (Zhang et al. 2001;Taki et al. 2010;Cheng et al. 2014;Wang et al. 2015;Zhou et al. 2015;Tamura and Yamane 2017;Trentini et al. 2017). However, it is unclear on the effect of thinning on both Chinese fir growth and soil physiochemical properties. In addition, less information is available on the effect of thinning on the change of tree growth in successive years of Chinese fir. How long will the effect of a thinning last? To improve the economic and ecological benefits of Chinese fir plantations and promote sustainable forestland management, in this study, the effects of thinning intensity on stand growth and soil physiochemical properties and their correlation were investigated by establishing sample plots in a nine-year-old Chinese fir forest (Huarong County, Hunan Province, China) and subjecting them to thinning of different intensities. With the purpose of identifying an appropriate thinning intensity for Chinese fir plantations and providing theoretical criteria for the sustainable operation of Chinese fir forests.

Study site
Sample plots were established in the Tashi State-Owned Forest Farm (29°10"29"" − 29°48"0"'N,112°18"0'' − 113°0"0'" E), which is located in Huarong County, Yueyang City, Hunan Province, China, and has an average altitude of 350 m. The study area has a subtropical moist land monsoon climate. The temperate climate is characterized by four distinct seasons, abundant heat, concentrated rainfall, a short cold winter, and a long hot summer. The study area has an annual average air temperature of 16.5°C, annual accumulated temperature of 6074°C, multiple-year average frost-free period of 264 d, and average annual precipitation of 1270 mm. The soil mainly consists of laterite. The dominant tree species in the region

Plot settings and survey
The thinning experiments were initially established in November of 2014 when the Chinese firs were nine years old. In November 2014, 12 stands measuring 20 × 30 m in area with uniform tree distribution and consistent site conditions (topographical factors and soil type) were selected as sample plots, which had an initial planting density of approximately 2500 stems ha −1 . The sample plots were subjected to thinning of four different intensities (three samples plots for each treatment): T1 (0%, 2500 stems ha −1 ), T2 (20%, 2010 stems ha −1 ), T3 (30%, 1750 stems ha −1 ), and T4 (40%, 1500 stems ha −1 ). Before thinning, tree scaling (tree height H [m] and diameter at breast height D [cm]), location, and numbering were conducted to each of the sample plots. The basic information of sample plots was shown in Table 1.
The sample plots were surveyed after the sample plot configuration. The survey data were conducted during the dormancy period of each subsequent year. The position for DBH measurement was marked during the initial survey, thereby ensuring DBH was measured at the same position during subsequent annual surveys. The DBH of trees with a DBH>5 cm was measured. Tree height was calculated using the heightdiameter curve method. At least 20 trees were sampled in each sample plot, and the height of three to five trees in each diameter class was measured using a fishing pole or a HG18CGQ-1 altimeter. The timber volume of Chinese fir stem was calculated according to the formula (Duan et al. 2016): where V is the timber volume of an individual Chinese fir tree (m 3 ), D is the tree DBH (cm), and H is the tree height (m). The stand volume of a sample plot was calculated as the sum of the timber volumes of individual trees and was expressed as the timber volume per hectare of stand.

Soil sampling and measurements
Soil samples were taken in November 2014 and 2019. The soil in each sample plots was sampled using the 5-point sampling method. Three samples were collected from each of the two topsoil layers: 0-20 cm and 20-40 cm. The samples were collected using a 100 cm 3 cutting ring and placed in aluminum boxes for testing their physical properties. Then, the samples from each topsoil layer were mixed even and placed in ziplock bags. The mixed soil samples were air-dried, ground and screened using a sieve (100 openings per square inch), and tested in a laboratory for their chemical properties. Soil physical properties, such as bulk density (g cm −3 ), capillary porosity (%), air-filled porosity (%), and total porosity (%), were tested using the methods described by Zhang et al. (2020). Organic matter content was determined using the potassium dichromate oxidation method (Liu et al. 2012). Total nitrogen was determined using an elemental analyzer. The soil available nitrogen content was determined using an alkali solution diffusion method. The total phosphorus content and available phosphorus contents were determined using a spectrophotometer (UV-1900, Japan). Total potassium and available potassium contents were measured using alkaline fusion-flame photometry with a Flame Photometer (CANY Instrument, Shanghai, China). Total nitrogen, available nitrogen, total phosphorus, available phosphorus, total potassium, and available potassium contents were determined using the methods described by Cheng et al. (2017). In order to measure the soil pH values, the soil:water suspensions (1:5) were stirred for 30 min and then measured by a pH meter (LEICI, China). For each plot, the average value of each parameter for each soil depth layer was calculated. The change of soil physical and chemical properties was analyzed by calculating the ratios of values after thinning (Zhang et al. 2019) to those before thinning .

Data processing
A one-way analysis of variance (ANOVA) method was adopted to access the significant differences among the thinning intensities (four level) on the tree growth (height, DBH, individual volume and stand volume), as well as the soil properties. Significant differences were measured by one-way ANOVA followed by a Tukey test, and significance was determined at the 0.05 level. The Shapiro-Wilk test was performed to the diameter class data to check their normality of distribution. These statistical analyses were carried out on SPSS25.0 software (IBM SPSS statistics) and the figures were mapped on Origin 2021 software (Origin Lab Corp., Northampton, MA, U.S.A). Redundancy analysis was used to determine the relationship between the soil physicochemical properties after five years of thinning and tree growth on Canoco 5.0 software (Microcomputer Power, Inc., Ithaca, NY, U.S.A). Figure 1 shows the current annual increments of tree height (a), DBH (b), individual volume (c) and stand volume (d) under different thinning intensities of Chinese fir, respectively. As shown in Figure 1, the effect of thinning in promoting tree height, tree DBH, individual volume and stand volume increment diminished year by year. As shown in Figure 1(a), the annual height increments of T1, T2, T3, and T4 in 2019 were 24.2%, 25.0%, 25.7%, and 26.3%, respectively, of those in 2015. As shown in Figure 1(b), the annual DBH increments of T1, T2, T3, and T4 in 2019 being 52.6%, 55.6%, 56.3%, and 81.9%, respectively, of those in 2015. As shown in Figure 1(c), the annual individual volume increments of T1, T2, T3, and T4 in 2019 being 41.7%, 46.2%, 46.7%, and 47.1%, respectively, of those in 2015. As shown in Figure 1(d), the annual stand volume increments of T1, T2, T3, and T4 in 2019 being 70.6%, 74.0%, 73.8%, 63.9%, respectively, of those in 2015. Table 2 shows the growth changes of Chinese fir plantations with different thinning intensities. The average annual height increments of T2, T3, and T4 increased by 11.1%, 11.1%, and 27.8%, respectively, as compared to T1. In particular, the average annual height increments of T4 was significantly higher than that of T1 (P < 0.05). The average annual DBH increments of T2, T3, and T4 increased by 40.0%, 55.0%, and 87.5%, respectively, as compared to T1. A comparison of the average annual DBH increments of stands subjected to thinning of different intensities showed that the difference between T2 and T3 was nonsignificant (P > 0.05) but the difference between the other pairs of treatments was significant (P < 0.05). The average annual tree volume increments of T2, T3, and T4 increased by 12.5%, 37.5%, and 50.0%, respectively, as compared to T1. The difference between T1 and T3/T4 was significant (P < 0.05), whereas that between T1 and T2 and that between T3 and T4 were nonsignificant (P > 0.05). The stand volume and average annual increment of stand volume of T4 was significantly lower than that of T1(P < 0.05), while T3 has a maximum value.  Figure 2, the DBH of four treatments were concentrated mainly in the range of 14-16 cm. The proportion of trees with a DBH ≥ 20 cm of T1, T2, T3 and T4 were 2.2%, 3.3%, 6.7% and 10.0%, respectively. As shown in Figure 3, the DBH of T1 was concentrated mainly in the range of 14-16 cm, while those of T2, T3, and T4 was concentrated mainly in the range of 16-20 cm. After 5 years of thinning, T4 had the highest ratio of trees with a DBH ≥ 20 cm (57.3%), followed by T2 (44.9%), T3 (40.5%), and T1 (21.3%). The number of large-DBH trees under the T4 thinning intensity was larger than under other thinning treatments. With the increase of thinning intensity, the peak of diameter distribution of Chinese fir stepped into the larger diameter level. A normality test showed that the DBH distribution of T4 in 2019 was nonnormal (P < 0.05), whereas those under other thinning treatments were normal (P > 0.05) (Tables 3, 4).

Effect of thinning on soil physical properties
The changes of the soil physical properties are shown in Table S1, S2. As shown in Table S2, the ratios of the bulk density in the 0-20 cm and 20-40 cm soil layers decreased with increasing thinning intensity. T1 differed significantly (P < 0.05) from T4 but non-significantly (P > 0.05) from T2 and T3 in the 0-20 cm soil layer. However, there were no significant differences (P > 0.05) among four treatments for the 20-40 cm soil layer, indicating that a higher intensity of thinning will notably influence the bulk density of 0-20 cm topsoil but have no obvious influences 20-40 cm soil layer.
The capillary porosities and total porosities in the 0-20 cm soil layer were higher in the T4 than that in the T1 (P < 0.05), and no significant differences were found among the four treatment in the 20-40 cm soil layer (P > 0.05). The differences in air-filled porosity between T1, T2, T3, and T4 were nonsignificant under different soil layers (P > 0.05).

Effect of thinning on soil chemical properties
The changes of the soil chemical properties are shown in Table  S3, S4. As shown in Table S4, no significant differences are observed in the ratios of the total nitrogen, total phosphorus, total potassium and available phosphorus contents under different thinning intensities and soil layers (P > 0.05). The ratios of soil organic matter and available nitrogen contents in both the 0-20 cm soil layers followed: T4 > T3 > T2 > T1, and T3 > T4 > T2 > T1 for the 20-40 cm soil layers. T1 differed significantly (P < 0.05) in the organic matter content in the 0-20 cm soil layer than T2, T3 and T4, and no significant differences (P > 0.05) in the 20-40 cm soil layer among the four treatments. The available nitrogen contents in the 0-20 cm and 20-40 cm soil layers were significantly higher in the T2, T3 and T4 than those in T1 (P < 0.05). In addition, the available nitrogen contents in the 0-20 cm soil layer were significantly higher in T3 and T4 than in T2 (P < 0.05). Available potassium contents in the 0-20 cm soil layer were significantly lower in T1 than in the other three treatments, and no significant difference was observed between the four treatments in the 20-40 cm soil layer (P > 0.05). Soil pH showed no significant differences among the four treatments in the 0-20 cm and 20-40 cm soil layers (P > 0.05).

Relationship between the tree growth and soil physicochemical properties
Redundancy analysis was conducted with the increment of four growth indexes as response variables and soil physicochemical properties as explanatory variable (enviromental variables). The results showed that the first two axes of the RDA plot explained 99.50% and 0.48% (99.98% in total) of the annual growth variation (Figure 4). A Monte Carlo permutation test showed that the AN and AK were significantly correlated (p < 0.05) with the tree growth in all environmental variables considered. In particular, the AN was the first factor that contributed to 77.0% of total variation (Table 5). As demonstrated in Figure 4, except △H, △DBH, △V were negatively correlated with SD and △SV was negatively correlated with SD and TK, the four growth indexes were positively correlated with other soil factors.

Discussion
In the first five years after thinning, the average annual increments in tree height, DBH, and timber volume increased with thinning intensity. In particular, the annual tree height, DBH, and timber volume increments of T4 increased by 27.8%, 87.5%, and 50.0% (Table 2), respectively, as compared to T1. From the annual increments of Chinese fir after thinning, the effect of thinning in promoting tree height, DBH, volume and stand volume increments diminished with time ( Figure 1). These results are consistent with the results from the studies by Gong et al. (2015) and Zheng et al. (2020). This was because thinning reduced stand canopy density and increased the nutrient space for the retained trees. In addition, thinning removed undergrown trees and retained trees with better genetic quality and greater growth potential (Trentini et al. 2017), and the improvement in soil physical conditions and the enhancement of soil available nutrients content, promoted the growth of Cunninghamia lanceolata, especially the growth of DBH and individual volume, which also may be due to the differences in environmental conditions in thinning plots, compared to the control plots (Table 1). However, compared with the control treatment, heavy thinning significantly reduced stand volume (Table 2). Stand volume is affected simultaneously by stand density and tree DBH and height growth (Lu et al. 2020). Decreasing stand density promotes the growth and increases the volume of individual trees but reduce the number of per-unit-area living standing trees. Stand volume depends on the thinning-enabled greater volume increment of individual trees relative to the volume loss due to the reduced number of trees (Xu 2020). The increase of stand volume after thinning failed to compensate for the decrease of stand volume caused by the decrease of tree number, which was consistent with other studies (Zhang et al. 2005;Xu et al. 2014;Quan et al. 2020). In addition, stand volume is closely related to biomass and carbon sinks (Lu et al. 2020). Under moderate thinning treatment, the stand volume increment had maximum values (Figure 1(d) and Table 2). Therefore, from the perspective of improving stand volume and biomass, a moderate thinning intensity (approximately 1750 stems ha −1 ) could be considered for adoption. Thinning improved the diameter class distribution of Chinese fir stands. Compared to the control treatment, thinning moved the diameter class distribution rightward; that is, the ratio of the number of large-DBH trees tended to increase after thinning. T4 had the largest number of trees with a DBH greater than 20 cm, being 2.68 times that under the control treatment (Figure 3). This indicated that increasing thinning intensity to an appropriate level facilitated the cultivation of large-DBH Chinese fir trees. In a study on the effect of thinning on the tree growth of a Chinese fir plantation in Kaihua, Zhejiang, China, Xu et al. (2014) found that thinning increased the number of trees with a DBH above 20 cm. In a study of a man-made middle-age Chinese fir pure forest in Guangxi, China, Lu et al. (2020) found that thinning increased the ratio of the number of medium-and large-DBH trees. With economic development, the market demand for small-and medium-diameter timber decreased, while that for large-diameter timber increased. Late thinning of Chinese fir forests with a high initial planting density will decrease stand diameter class (Sheng 2018;Liu et al. 2021). Thus, from the perspectives of large-DBH tree cultivation and stand structure improvement, thinning intensity can be appropriately increased in silvicultural practice.
Forest soil is a major component of forest ecosystem. Operational measures and anthropogenic disturbances can significantly affect soil quality (Bai et al. 2017). Some studies have shown that soil physical properties, such as bulk density and soil hydraulic parameters, change with depth, following   -Dumroese et al. 2010;Zhao et al. 2014;Chen et al. 2014;Zhou et al. 2015). Tarpey et al. (2008) observed that soil bulk density and penetration resistance increased as thinning intensity increased, in a red pine (Pinus resinosa) stand growing on a sandy soil, over a 57-year period. Bravo-Oviedo et al. (2015) found that the bulk density did not change among treatments in a Scots pine (Pinus sylvestris) stand after 40 years under a thinning treatment. This study showed that thinning reduced the soil bulk density in the Chinese fir plantation and increased the overall capillary porosity, air-filled porosity, and total porosity. The effect was greater on the 0-20 cm soil layer than on the 20-40 cm soil layer. These results are consistent with the results of previous related studies (Zhang et al. 2001;Zhou et al. 2018). The increase of soil porosity in surface soils in the heavy thinning treatment is likely related to root death of the removed trees, and root growth of the understory vegetation (Zhu and Dong 2016;Cheng et al. 2017;Shu et al. 2021). Some studies found that the decomposition of plant roots can promote soil aeration and aggregation, and soil particles adhesion, which contribute to soil formation, development, and stability (Verboom and Pate 2006;Lambers et al. 2009;Song et al. 2012).
From 2014 to 2019, compared to Chinese fir stands without thinning, the variation of nutrient contents in each layers of stands subjected to thinning of different intensities increased by different degrees. In particular, the nutrient contents in the 0-20 cm soil layer of T4 increased significantly as compared to T1. On the one hand, this would be a result of the altered soil temperature, soil water content, and microbial activity after thinning (Cheng et al. 2017;Liu et al. 2019). On the other hand, in our previous studies, it was found that thinning could promote the development of understory vegetation (Zhao 2021). The improved development of the understory plants after thinning had led to an increasing proportion of rapidly decomposable litter, which could effectively supplement the available nutrient content (for example, SOM, AN and AK) input to the soil (Teste et al. 2012;Muscolo et al. 2015;Wang et al. 2019;Xu et al. 2020).
Further, redundancy analysis ( Figure 4) showed that AN and AK contributed more to tree growth, and SOM was positively correlated with △H and △DBH, which was consistent with previous research results (Springsteen et al. 2010;Zhang et al. 2021). At the same time, it was found that soil bulk density was negatively correlated with tree growth, which was consistent with the results of Wang et al. (2011). Overall, the effect of thinning on soil nutrient characteristics in different soil layers showed that the response of topsoil nutrient to stand density was more obvious (Table S4). With the increase of soil depth, nutrient content may change due to the difference of root distribution in different soil layers and the influence of leaching and mineralization of soil nutrients (He et al. 2017). However, Zhou et al. (2015) found that soil chemical properties tended to deteriorate with an increase of high intensity thinning, in a natural mixed Chinese fir forest in southeast China, 15 years post-thinning. These different results may be related to tree age, site condition, and thinning regime. The specific influencing mechanism needs to be further studied.
The effect of thinning on soil fertility and tree growth plays an important role in forest management, as soil fertility is related to forest growth and site productivity, which are synergistic and interactive (Powers 2012). Improvements in soil fertility and tree growth could be affected by various factors, including anthropogenic disturbances (such as thinning measures) and natural environmental conditions (such as site and climate conditions). The appropriate thinning intensity of Chinese fir could be different in different regions and different ages. Therefore, in order to objectively assess the effects of thinning intensities on the tree growth and soil properties, and provide more reliable long-term forest management Figure 4. Relationship between the tree growth and soil physicochemical properties. △DBH, △H,△V, △SV, SD, TOP, CP, SOM, TN, TP, TK, AN, AP, AK denote diameter at breast height increment, height increment, individual volume increment, stand volume increment, soil density, total porosity, capillary porosity, soil organic matter, soil total nitrogen, total phosphorus, total potassium, available nitrogen, available phosphorus, available potassium, respectively. Table 5. Conditional effects of soil properties on tree growth from the summary of forward selection in the redundancy analysis (RDA).

Variables
Explains ( measures, it is recommended that future studies should focus on long-term continuous observations and the analysis of any observed dynamic changes, which may occur over time.

Conclusions
The results of this study indicated that thinning had a potential effect on the growth and soil properties of Chinese fir plantation. The average tree height, DBH and volume increase of Chinese fir plantation in the heavy thinning treatment were significantly higher than those of the control, which was conducive to the cultivation of largediameter Chinese fir, and the price of large-diameter Chinese fir was higher. Meanwhile, heavy thinning could also improve the physical and chemical properties of soil, but significantly reduced stand volume. In addition, the tree growth was significantly correlated with the soil nutrients (for example, AN and AK). Based on these findings, the heavy thinning intensity (approximately 1500 stems ha −1 ) was the optimum for maintaining economic and ecological benefits. From the perspective of improving stand volume and biomass, a moderate thinning intensity (approximately 1750 stems ha −1 ) could be considered for adoption. Chinese fir is a fast-growing tree species. Although this is a shortperiod experimental study conducted for five consecutive years, we revealed the effects of thinning on growth performance and soil properties in a Chinese fir plantation. Forest ecosystems are characterized by interactions between the vegetation, soil, and microorganisms. Thinning serves to accelerate the interaction processes. Observation of the effect of thinning on forest ecosystem is a complex, long-term process, and factors, such as forest age and site conditions may affect the outcome of tending. All this requires further observational studies.