Introduction

China is undergoing a shift in its forestry strategy from timber production to ecological rehabilitation (Zhou 2004). Plantations play an important role in both timber production and ecological protection because China has the largest and most extensive tree plantations in the world. The total plantation area in China is estimated at approximately 62 million ha representing 23.5 % of the global area given to tree plantations (SFA 2010; FAO 2011). In order to satisfy the high demand for timber products, the Chinese government planted fast-growing wood species (e.g., Chinese fir, Eucalyptus) over large areas. However, most timber plantations have not been efficiently managed and are monocultures with age distributions dominated primarily by young and middle-aged trees with a poorly developed understory. As a result, many serious ecological problems have emerged, such as low species composition, reduced biological diversity, and loss of water and nutrients (Ma et al. 2002a). These challenges have in turn resulted in reduced forest productivity and inadequate timber yields (Ares et al. 2010; McFarlane et al. 2010; Ma et al. 2002b), indicating the need for a renewed emphasis on forest management in China, now and in the future.

Tree thinning is the most widely applied silvicultural practice worldwide in plantations as well as in natural forests. Thinning increases the growth of residual trees, forest yield, and timber revenue (Mäkinen and Isomäki 2004a, b; Juodvalkis et al. 2005; Zeide 2001, 2004). Many studies have reported that thinning benefits growth of the remaining trees by reducing competition for light, water, and nutrient resources (Cañellas et al. 2004). However, it has also been argued that while thinning may increase the diameter and volume of individual trees, it does not significantly affect tree height or stand volume (Pan et al. 2003; Mäkinen and Isomäki 2004c; Cañellas et al. 2004). For example, Mäkinen and Isomäki (2004c) reported that the average incremental growth in tree diameter increased with thinning intensity, but the annual increment in dominant tree height was not affected by thinning in Norway spruce (Picea abis (L.) Karst) over a measurement period averaging 27 years. Baleshta et al. (2015) also found that thinning did not affect the height of the remaining trees. Cañellas et al. (2004) found that the total basal area and stand yields were not significantly affected by thinning management.

However, novel management strategies have stimulated more interest in thinning, including its effects on forest structure and biodiversity (Parrotta et al. 2002; Montes et al. 2005a, b; McElhinny et al. 2005), improvement of wildlife habitat (Hagar et al. 1996; Verschuyl et al. 2011), acceleration of the functional diversity of the understory (Strahan et al. 2015; Nunes et al. 2014; Ares et al. 2010), and decreasing fire hazard (Johnson et al. 2007). Understory vegetation contributes to a wide variety of ecosystem functions by improving floristic biodiversity, limiting soil erosion and nutrient losses due to leaching, and facilitating nutrient cycling (Nilsson and Wardle 2005; Wayman and North 2007). Thinning can alter the understory indirectly by changing canopy densities and forest microclimate parameters, e.g., light, soil water, and nutrient availability. However, the effects on biodiversity of the thinning of fast-growing species are a subject of recent debate (Tang et al. 2007). Some reports have argued that thinning can increase biodiversity, for example, by increasing understory vegetation (Taki et al. 2010; Chan et al. 2006); while others have concluded that thinning either decreases understory diversity (Nagai and Yoshida 2006; Metlen and Fiedler 2006) or has no influence (Lei et al. 2007; Gilliam et al. 1995). These contradictory reports on the effects of thinning management on the biodiversity of the understory may be attributed to variations in pre-treatment site conditions (Fulé et al. 2005), differences in thinning intensity and timing (Ares et al. 2010), understory dynamics (He and Barclay 2000; Thomas et al. 2008), and/or stand age (Juodvalkis et al. 2005). Consequently, regional studies are required for assessing the impacts of thinning that compare different silvicultural practices in similar geographical environments. The effects of the understory on soil fertility and overstory-understory interactions following thinning also needs to be investigated in terrestrial forest ecosystems, especially in fast-growing forest and monoculture ecosystems (Royo and Carson 2006).

Chinese fir (Cunninghamia lanceolata (Lamb.) Hook.), a fast-growing coniferous evergreen tree with high yield and excellent wood quality, is the most important timber species in China. The earliest plantings of Chinese fir can be dated back 1000 years (Yu 1997). The planted area is estimated to be over 11.27 million hectares, accounting for approximately 16.3 % of all tree plantations in China and 4.3 % of trees in plantations worldwide (SFA 2010; FAO 2011). However, since the 1980s, while the planting area has more than doubled, timber yields have declined over successive rotations because of ecological deterioration, defined by physical, chemical, and biochemical parameters, and nutrient loss from soil caused by short rotation cycles, clear-cutting, and controlled burning (Ma et al. 2007; Tian et al. 2011a).

Thinning has been regarded as an effective method for managing these problems in Chinese fir plantations and can be dated back to the 1960s (Tian 2005). During a complete rotation of Chinese fir, thinning is conducted 2–3 times, with the first thinning usually done at 7–9 years and the second at 15–16 years (Yu 1997). Previous research has investigated the effects of thinning on understory diversity and biomass (Zhao et al. 2012; Cheng et al. 2014), tree growth (Zhao et al. 2012), soil physical and chemical characteristics (Zhang et al. 2001), and the physical and mechanical properties of wood (Deng et al. 2009) in Chinese fir plantations. However, most research has focused on the effects of thinning on either tree growth or understory vegetation over short periods; the longer-term relationships among tree growth, understory vegetation, and ecophysiological responses in the soil environment after thinning are not clear. The effects of tree thinning on understory diversity and community structure, soil properties, and ultimate consequences for tree growth and stand yields need to be resolved (Chan et al. 2006). The multiple-use forest management (MFM) approach has diverse objectives such as providing both timber and non-timber forest products and environmental services (García-Fernández et al. 2008). According to the multiple-use concept, the dynamics of understory vegetation, effects on soil nutrient, and forest growth after thinning in Chinese fir plantations need to be explored.

In view of this, we hypothesized that thinning management would trigger dynamic effects on understory vegetation at different developmental stages of within a rotation of trees within a Chinese fir plantation. We also hypothesized that thinning management would increase understory diversity and biomass, decrease soil erosion, and increase soil nutrient availability, which may benefit Chinese fir growth over the course of a full rotation. Therefore, we investigated (1) the growth dynamics of monoculture Chinese fir plantations during the young (after the first thinning, 9 years old) and middle stages (after the second thinning, 16 years old) during a typical rotation; (2) the dynamics of understory diversity and biomass as function of thinning management; and (3) the dynamics of soil physical and chemical properties as a function of thinning. We also analyzed the mutual-relationship between understory effects and Chinese fir growth in order to provide guidance for long-term management practices of Chinese fir plantations.

Materials and methods

Study site and experimental design

The two study sites (Dangan for the young stage, Pushang for the middle stage) are located at the Pushang National Forest Farm, Shunchang county, Nanping city, in Central Fujian Province, Southern China. The two sites are situated about 6.6 km apart (Dagan site: 26° 57′ 57.8″ N, 117° 43′ 41.9″ E; Pushang site: 26° 56′ 20.1″ N, 117° 47′ 8.8″ E) (Fig. 1). The region is characterized by a subtropical monsoon climate, mean annual average temperature of 18.5 °C, and mean annual average rainfall of 1621 mm. The average annual temperature and precipitation during study period (2004–2011) is shown in Fig. 2. The rainy season is concentrated in spring and summer. The average annual relative humidity is 83 %, annual average evaporation is 1413 mm, and the average frost-free season is 305 days from early March to November. Elevations range from approximately 200 to 300 m. The soil is classified as mountainous red clay loam covered with thin humus and acidic oxisols, which was developed on parent shale. The impeded soil layer was more than 100-cm deep.

Fig. 1
figure 1

Locations of the study sites (Dagan and Pushang) in Shun Chang Country, Nanping City, Fujiang province, China

Full size image
Fig. 2
figure 2

Mean annual precipitation (vertical bars) and mean annual temperature (curves) of the study area during 2004–2011

Full size image

Stands used in this study were originally broad-leaved forests and now were dominated by Chinese fir plantations planted by clear-cut and burning practices. Two monoculture plantations of Chinese fir including stands in the young and middle stages located in Dagan and Pushang town, respectively, were selected. The total area was 22.9 ha, including 11.2 ha at Dagan and 11.7 ha for Pushang. The young stage stands were planted in 1995 and the middle stage stands were planted in 1988. Thinning experiments were established at the Pushang National Forest Farm in February in 2004 when the young stage trees were 9 years old and the middle stage trees were 16 years old. The young and middle stages of Chinese fir plantations were thinned for the first and second time, respectively. For the middle stage, the first thinning time was 1996 and the average planting density was 3210 stems per hectare with an average thinning intensity of 48.6 %. Characteristics of the plantations prior to and post-thinning are presented in Table 1. According to the local guidelines on thinning management for conifers proposed for Fujian province (Ma et al. 2007), the average thinning intensities were 51.8 and 30.1 % of the stem numbers for the young and middle stages, respectively. For the thinned stands, prior to thinning, the average planting densities were 3058 and 1450 stems ha−1 and the average residual densities were 1458 and 1000 stems ha−1, for young and middle stage, respectively; for unthinned stands, the average planting densities were 3042 and 1417 stems ha−1, respectively. The stand volume were 217.18 and 189.53 m3 ha−1 in young and middle stage, respectively, prior to thinning, while the residual stand volume were 137.18 and 149.03 m3 ha−1 in young and middle stage, respectively, after thinning. The coverage changes were 0.98 and 0.75, respectively, in the young stage, and 0.85 and 0.80, respectively, in the middle stage before and after thinning. Prior to thinning, for the young stage, the average diameter at breast height (DBH) and height for unthinned and thinned stands were 12.76 cm and 9.26 m and 13.32 cm and 9.86 m, respectively; for the middle stage, the average DBH and height was 15.83 cm and 12.85 m and 16.41 cm and 12.36 m, respectively, for unthinned and thinned stands. In thinned plots, trees that were small or overcrowded or suffered from diseases and pests were cut and removed, leaving large and healthy individuals. Control areas (unthinned stands) were not treated other than actively protecting the understory vegetation. In mid-April 2004, we set six “large” sample plots of 20 m × 20 m in thinned and unthinned areas at three different elevations at the young and middle stages of a rotation. In the same large plot, the distance between thinned and unthinned plots at the same elevation from one another was almost 80 m. Thus, a total of 12 large plots were sampled. To minimize edge effects, all large plots were placed at least 100 m away from each other.

Table 1 The general configuration among young and middle stage in Chinese fir plantations before and after thinning
Full size table

Sampling collection

Floristic surveys were carried out in each large plot, and trees were labeled and the tree densities (numbers of trees per hectare) recorded. The DBH was measured with a tape meter. The height of individual trees was measured using a laser altimeter (Haglof Vertex IV, Sweden). The growth response of Chinese fir trees to thinning was tracked in 2004, 2005, 2006, 2007, and 2011.

Floristic surveys of understory vegetation, differentiated as shrub and herbaceous layers, were carried out in five “small” 5 × 5 m subplots for shrub species and 1 × 1 m subplots for herb species within each large plot (one subplot in the center and one each on four corners surrounding the central subplot). We categorized understory vegetation as shrubs (height <5 m) and herbs (height <0.5 m with non-woody parts) (Zhang and Zhang 2007). We combined data from the five small subplots to characterize each large plot. Within understory layers (shrub and herb), the species were categorized and enumerated according to information listed on the eFlora of China website (www.eflora.cn). The coverage, frequency, height, and the basal diameter for shrubs were also recorded. These data were used to calculate relative abundance (density), relative frequency, relative coverage, and relative dominance for quantitative analysis. In order to calculate the understory biomass, we collected all the individuals of each species in each small subplot (also divided into shrub and herb layers) and immediately weighed the fresh weight including branches and leaves, then took them to the laboratory. All the samples were oven-dried at 80 °C for 48 h and weighed. The total biomass for each understory layer from each small subplot was used to calculate the biomass per hectare. Understory parameters were measured in 2004, 2005, 2006, 2007, and 2011.

Soil samples were collected in April in 2005, 2007, and 2011. To measure bulk density and soil water content, five intact soil cores were taken randomly from mineral soil at 0–10-cm depth in each large plot (20 × 20 m) using a soil wreath knife (200 cm3). The soil physical properties were measured according to Li and Shao (2006). Soil bulk densities (D BD) were determined by mass of soil per unit volume (sum of solids and pore space) and the soil cores were oven-dried at 105–110 °C. Each soil core was put into a salver to absorb water via filter paper until it reached a steady weight, usually 12 h later. Then, the maximum water-holding capacity (Whc) was measured. Then, the soil cores were put on a sand salver and allowed to drain for 2 h in order to calculate soil capillary water content (Wc). Soil capillary porosity (Scp) was calculated according to the equation Scp = 0.1 × Wc × ds. Here, the Scp is the soil capillary porosity (%), Wc is the soil capillary water content (g kg−1), and ds is the soil density (mg m−3). Soil non-capillary porosity (Sncp) was calculated according to the equation Sncp = 0.1 × (Whc − Wc) × ds. Soil total porosity was the sum of Scp and Sncp.

Five additional soil samples were collected from the 0–10-cm soil layer to determine soil chemical properties. They were air-dried, sieved (less than 2 mm), and stored at 4 °C for chemical analysis. Before sieving, all debris and plant remains were removed. Ten-gram aliquots of dry soil sample were extracted with 25-mL CO2-free distilled water for 30 min (stirring at 0, 5, and 10 min); then, the pH was measured with a precision pH meter (PHS-2, China). Soil organic matter was measured using the potassium dichromate oxidation-external heating method (Zhang et al. 2009). Soil hydrolysable N concentration (g kg−1) was measured by a semi-micro Kjeldahl method using a KDN-04B Nitrogen Analyzer (TOP Instrument, Zhejiang, China) (Sims et al. 1995). Soil available P (g kg−1) was measured using the molybdenum blue colorimetric method with a UV1800PC spectrophotometer (AOXI Instrument, Shanghai, China) (Institute of Soil Academia Sinica 1978). Soil available K was measured by alkaline fusion-flame photometry using an AP1200 Flame Photometer (CANY Instrument, Shanghai, China). All these determinations were carried out at the Forestry College Analytical Laboratory in Fujian Agriculture and Forestry University.

Data analysis

Tree growth at individual and stand levels

To evaluate the effects of thinning on tree growth, individual tree and stand volumes were calculated. Individual tree volume was calculated by the formula V = 0.00005806D 1.955485 H 0.894033 (Hong and Lin 1993). D is the diameter at breast height (cm) and H is the tree height (m). The total volume was calculated as the sum of all trees in the same plot. The periodic annual increment (PAI) values for individual trees and stands in each plot were calculated by Pressler’s formula (Philip 1994). Each measurement period was based on the values of living trees that had been since the beginning of the study. The formula was as follows:

$$ {P}_n=\frac{y_a-{y}_{a-n}}{y_a+{y}_{a-n}}\frac{200}{n} $$

where P n is PAI, n is the number of years between inventories, y a is the measure (quadratic mean diameter, basal area, or volume) in inventory a, and y a-n is the measure in inventory years a–n. Quadratic mean diameter (QMD) was calculated as the diameter of a tree with average basal area at breast height. Stand basal area per hectare was calculated by summing the values for all trees in the plot. Volume was calculated at both the tree and stand levels. The stand volume per hectare was determined by counting the sum of all trees in the plot.

Understory diversity, biomass, and soil properties

We calculated the relative abundance (density expressed as individuals ha−1), frequency, dominance, and importance value of each species across all large plots at all elevational ranges. The importance value index (IVI) was used to describe the species composition of the plots. For shrubs, the IVI of a species was calculated by the average of its relative dominance, relative density, and relative frequency, while for the herbaceous layer, the IVI value was calculated by the average of its relative density, relative coverage, and relative frequency, because it was difficult to determine the basal area for herbs. The parameters were in the following:

  1. (1)

    Relative dominance = total basal area for a species/total basal area for all species × 100

  2. (2)

    Relative frequency = frequency of a species/sum of frequencies for all species × 100

  3. (3)

    Relative coverage = coverage of a species/sum coverage of all species × 100

  4. (4)

    Relative density = number of individuals of a species/total number of individuals × 100

The theoretical range for relative dominance, relative density, relative coverage, and relative frequency was 0–100 %, so the IVI of a species could vary between 0 and 100 %, because each IVI was the average of three percentages.

To evaluate understory species diversity, three indices were chosen according to Pielou (1969):

  1. (1)

    Richness (S): S is the number of species within a plot,

  2. (2)

    Shannon-Wiener’s diversity index (H′): \( H\hbox{'}=-{\displaystyle \sum_{i=1}^S{P}_i} \ln {P}_i \),

  3. (3)

    Simpson diversity index (D’): \( {D}^{\prime }=1-{\displaystyle \sum_{i=1}^S{P}_i^2} \),

  4. (4)

    Pielou Evenness index (J): J = H′/ln (S);

Where P i is the proportion of individuals belonging to species i and S is species richness represented by the total numbers of species in each large plot.

The Shannon-Wiener diversity index (H′) was used as a synthetic measure of community structure, because it reduces the effects of rare species (Pielou 1975). The Pielou Evenness index (J) was selected to quantify community similarity.

To evaluate the effects of thinning treatments on the shrub and herb layer, the diversity and evenness indices were analyzed by two-way ANOVA with treatments (thinned and unthinned) and time after thinning (1–7 years) as the principal factors. Mean comparisons were performed by the least significant difference test (p < 0.05). To evaluate the differences of understory diversity, biomass, soil physical and chemical properties among thinned and unthinned stands during the thinning period, statistical differences between means were determined using independent t tests and Tukey’s Multiple Range test (p < 0.05). All statistical analyses used the SPSS Statistical Package (SPSS 12.0, SPSS Ins., IL, U.S.A).

Results

Quadratic mean diameter, basal area, and volume increment at tree and stand level

The effects of thinning on Chinese fir growth over the 7-year response period are shown in Fig. 3. Panels a and c illustrate that the QMD and volume greatly increased in thinned stands as compared to that unthinned stands in both young and middle stages of the rotation. During the monitoring years, the QMD values in unthinned and thinned stands increased by 12.4 and 19.7 % at the young stage and 19.8 and 20.0 % in the middle stage, respectively. Individual volumes increased by 47.3 and 77.7 % at the young stage and 43.0 and 63.0 % in the middle stage, respectively. Panels b and d in Fig. 3 indicate that both the basal area and stand volume were higher in unthinned stands than in thinned stands; however, the increments were all higher in the thinned stands. Specifically, the basal area of trees from the unthinned and thinned stands increased by 9.1 vs. 28.7 % during the young stage and 19.4 vs. 39.7 % during the middle stage, respectively. Stand volumes for unthinned and thinned stands increased 26.9 and 58.4 % in the young stage and 33.6 and 61.7 % in the middle stage of the rotation, respectively. According to Pressler’s formula, the periodic annual increment (PAI) of the QMD, individual volume, stand basal area, and stand volume were all higher in thinned stands than in unthinned stands, and the PAI of individual volume and stand volume were significantly higher in thinned stand in the young and middle stages of the rotation (Table 2).

Fig. 3
figure 3

Mean of QMD (a), stand basal area (b), individual volume (c), and stand volume (d) among thinned and unthinned stands from 2004 to 2011 in young and middle stages of Chinese fir plantations. Values are mean ± stand error

Full size image
Table 2 Periodic annual increment (PAI) of quadratic mean diameter (QMD), individual volume, stand basal area, and stand volume among thinned and unthinned stands from 2004 to 2011 in young and middle stages of Chinese fir plantations
Full size table

The two-way ANOVA for all data in different measurement years illustrated that treatment and time had significant effects on the PAI of the QMD, individual volume, stand basal area, and stand volume of trees in the young stage of the rotation (Table 3). However, in the middle stage, the PAI of QMD and individual volume significantly varied among thinned and unthinned stands at all measurement periods. The trees’ growth after thinning was seldom influenced by treatment × time interactions except for the PAI of individual volume in both young and middle stands.

Table 3 Effects of thinning treatment (thinned and unthinned) and year (1–7 years after thinning) on the periodic annual increment (PAI) of quadratic mean diameter (QMD), stand basal area, individual volume, and stand volume in a two-way ANOVA analysis with a homogeneity of variance test, followed by the least significant difference test
Full size table

Understory vegetation 1–7 years after thinning

Floristic composition and IVI

There were more species in the understory of thinned than unthinned plots in both young and middle stage stands (Table 4). Seven years after thinning, a total of 35 families, 43 genera, and 54 species were recorded in 1200 m2 within young stage plots, compared to 14 families, 15 genera, and 15 species within unthinned plots; a total of 49 families, 74 genera, and 93 species were sampled within 1200 m2 in middle stage, compared to 40 families, 60 genera, and 73 species in unthinned plots. It was concluded that species composition increased by 35 species in the young stage thinned plots while species diversity did not increase in the unthinned plots. In the middle stage, species increased by 40 in thinned plots and by 27 in unthinned plots. We found that species diversity increased the most during the first 2 years after thinning but fewer new species became established at later times post-thinning.

Table 4 Summary of floristic composition of understory communities among thinned and unthinned stands in young and middle stage 1 to 7 years after thinning
Full size table

The five most abundant species in each layer 7 years following thinning (2011) are shown in Table 5. In the young stage, due to high canopy density and low light transmittance in unthinned stands, the dominant species were mainly shade-tolerant shrub species such as Callicarpa kochiana, Ilex asprella, and Ilex pubescens and fern species such as Allantodia metteniana, Dryopteris fuscipes, and Parathelypteris glanduligera. In the thinned stands, sun-tolerant species like Styrax faberi, Callicarpa formosana, and Vaccinium carlesii were dominant in the shrub layer, and Lophatherum gracile, Gahnia tristis, and Woodwardia japonica were primary species in the herb layer (Table 5). In the middle stage plots, the dominant species in the shrub layer were similar between the thinned and unthinned stands, but the light-demanding species Maesa japonica, C. formosana, and S. faberi became dominant in the thinned stand. The herb layer was also occupied by fern species both in thinned and unthinned stands in middle stage.

Table 5 The top five most abundant species in different understory vegetation layer, according to decreasing order of importance value index (IVI) in thinned and unthinned plots in young and middle stage 7 years after thinning. Numbers in brackets represent IVI
Full size table

Species diversity

We analyzed the species diversity for the shrub and herb layers separately during the monitoring period (Tables 6 and 7). Shannon-Wiener diversity (H′) and Simpson diversity (D’) indices were higher in the shrub than the herb layers, while the two diversity indices showed a trend towards increasing diversity in the herb layer during the 7 years after thinning in both thinned and unthinned stands. Statistical comparisons of the Shannon-Wiener diversity (H′), Simpson diversity (Dʹ), and Pielou Evenness index (J) showed significantly higher species diversity 1 and 2 years after thinning, than those in the unthinned plots of the same age in the young stage (Table 6). However, there were no significant differences among thinned and unthinned stands for the three diversity indices in middle stage except Shannon-Wiener diversity 7 years after thinning (Table 7).

Table 6 Independent t test analyzing Shannon-Wiener (H′), Simpson diversity (D’), and Pielou Evenness index (J) indices of shrub and herb layer between different thinning treatments (thinned and unthinned) among young stages of Chinese fir plantations 1–7 years after thinning
Full size table
Table 7 Independent t test analyzing Shannon-Wiener (H′), Simpson diversity (Dʹ), and Pielou Evenness index (J) indices of shrub and herb layer between different thinning treatments (thinned and unthinned) among middle stages of Chinese fir plantations 1–7 years after thinning
Full size table

Two-way ANOVA analysis indicated that understory biodiversity changed significantly as a function of time (Table 8). For the young stage, thinning management had significant effects on Shannon-Wiener diversity (H′) and Simpson diversity (D’) in both shrub and herb layer; however, for the middle stage, thinning significantly affected both indices in the herb layer. The Pielou Evenness index (J) was only significantly affected by thinning in the middle-stage shrub layer. There were no significant treatment × time interaction effects on understory biodiversity in either the young or middle stages of the rotation except for Pielou Evenness index (J) in the young-stage herb layer.

Table 8 Effects of thinning treatment (thinned and unthinned) and year (1 to 7 years after thinning) on the Shannon-Wiener index, Simpson diversity index, and Pielou Evenness index in shrub and herb layer among different developmental stages in a two-way ANOVA analysis with a homogeneity of variance test, followed by the least significant difference test
Full size table

Understory biomass

Understory biomass was significantly higher in both the shrub and herb layers in the thinned stands than in the unthinned stands (Fig. 4). Seven years after thinning, the biomass in the shrub and herb layer in the thinned stand had increased by 771.87 kg ha−1 as compared to that in the young unthinned stands, and increased by 615.36 kg ha−1 in the middle stage (Fig. 4c, f). Seven years after thinning, there was a greater increment increase in the biomass of the shrub layer than in the herb layer between thinned and unthinned stands (Fig. 4a, b, d, e). Specifically in the young stage, understory biomass increased 4400.97 and 299.55 % in the shrub and herb layers, respectively, in thinned stands as compared to that in unthinned stands. However, the increment increase was 781.66 and 22.07 % in the shrub and herb layers, respectively, in the middle stage.

Fig. 4
figure 4

Biomass of understory vegetation in the shrub and herb layer among thinned and unthinned stands in young and middle stage of Chinese fir plantations (kg ha−1). **Significant differences between thinned and unthinned stands in the same layer by independent t test at p = 0.01

Full size image

Soil physical and chemical properties

The effects of thinning on soil physical properties 1, 3, and 7 years following thinning are shown in Table 9. Soil physical properties were more affected by thinning in the young stage. In the young stage, the bulk soil density decreased with time following thinning, and was significantly lower in thinned stands than in unthinned stands (p < 0.05). The maximum soil Whc increased with time following thinning and was significantly higher in the thinned stand than in the unthinned stand during the young stage (p < 0.05). The soil capillary porosity and total porosity during the 7 years following thinning increased by 12.6 and 5.0 %, respectively, in the young stage plots. The soil non-capillary porosity decreased significantly with time in the thinned stand and trended lower in the unthinned stand among both young and middle stage stands.

Table 9 Independent t test analyzing soil physical properties in the surface layer (0–20 cm) between different thinning treatments (thinned and unthinned) among young and middle stages of Chinese fir plantations 1–7 years after thinning
Full size table

Soil pH was higher in thinned stands than in unthinned stands in both young and middle stages (Table 10). Soil organic matter increased with increasing time post-thinning, and became significantly higher in the 7 years following thinning, than that in unthinned stands (p < 0.05). The same trend was also found in soil available P and hydrolysable N. For example, after thinning 7 years, the available soil P and hydrolysable N increased by 37.6 and 18.2 %, respectively, in the young thinned plots over unthinned plots, and increased by 50.1 and 13.3 %, respectively, in the middle stage thinned over unthinned plots.

Table 10 Independent t test analyzing soil chemical properties in the surface layer (0–20 cm) between different thinning treatments (thinned and unthinned) among young and middle stages of Chinese fir plantations 1–7 years after thinning
Full size table

Relationships among tree growth and understory biomass

The linear model showed that increasing understory biomass was not associated with reduced tree growth at the tree and stand levels (Fig. 5). In fact, there was a positive correlation between understory biomass and incremental increase in tree volume, which reached statistically significant levels at the individual tree level and was nevertheless observable at the stand level (p < 0.05). This was true for both the herb layer and entire layer, indicating that understory vegetation development could improve stand soil properties and increase individual tree growth.

Fig. 5
figure 5

The linear relationship between biomass of understory vegetation with individual tree volume and stand volume increment was calculated 1–7 years after thinning, combining both thinned and unthinned stands (n = 22). R 2 was the Pearson’s linear correlation coefficient at p = 0.05 (two-tailed)

Full size image

Discussion

Tree growth at the individual and stand level

Our experiments have shown that thinning increases both the diameter and volume increments of trees at the individual level. Thinning promoted the PAI of both QMD and tree volume, especially in young stands, relative to middle stage stands (Figures a, c; Table 2), which is in agreement with other studies (Juodvalkis et al. 2005; Cañellas et al. 2004; Mäkinen and Isomäki 2004a, b; Varmola and Salminen 2004). For example, Juodvalkis et al. (2005) reported that tree growth responded to thinning depending on the age of the stand. They reported that thinning had the greatest effect in trees aged 10–20 years but that effect decreased sharply in older stands and further increases in volume increment were unachievable in 50–60-year-old stands. Mäkinen and Isomäki (2004a) reported that thinning did not accelerate the growth of the largest trees in a stand of Scots pine. The growth of young and small-sized trees was enhanced after thinning in terms of stand crown increment compared to older stands (Juodvalkis et al. 2005; Varmola and Salminen 2004). In our study, tree volume increased as a result of thinning to a greater degree in young stage plots than the middle stage at both the individual and stand level (Tables 2 and 3), probably because the initial stand densities were higher before thinning and the extent of thinning was greater in the young stage (Simard et al. 2004). However, contradictory results have been obtained with respect to how the growth of individual trees at different ages responds to thinning. Cañellas et al. (2004) found that the growth of large stems was stimulated by thinning more than that of smaller trees, suggesting that large trees may have a greater capacity for resource acquisition, allowing them to take advantage of the increased resource availability after thinning.

In our study, the absolute basal area and volume per hectare were lower in thinned stands than in unthinned stands, but the PAI of the basal area and stand volume were higher in thinned stands at both young and middle stages of the Chinese fir rotation (Table 2). This is in accordance with Simard et al. (2004), who reported that total stand volume per plot was lower in thinned than control stands but that the increment of increase for both mean stand diameter and height had increased during 5 years following thinning in paper birch (Betula papyrifera Marsh.) stands. However, our study found that the increased volume of individual trees following thinning may compensate for the reduction in stand yield and predicted stand volume and therefore, thinned stands may catch up to unthinned stands over time (Fig. 1d). These results were consistent with those of Rytter and Stener (2005), who reported that the faster growth that occurs with heavier thinning might compensate for the loss in total volume production. Whether the stand volume in thinned stands exceeds that in unthinned stands over time can only be determined with long-term measurements. In our study, treatment and time were significant factors in the growth of Chinese fir stands at both the individual and stand level, even at the young stage (Table 3), suggesting that thinning management had profound effects due to a higher growth response potential at the individual level, contributing to an increase in basal area.

Understory diversity and biomass changes from 1 to 7 years after thinning

Whether fast-growing monoculture species (e.g., Eucalyptus and Chinese fir) exhaust resources and cause lower biodiversity has been a topic of international debate. Our results confirmed our first hypothesis that understory diversity could be accelerated by thinning management in Chinese fir monocultures (Tables 4 and 5), consistent with the studies of many others (Wang et al. 2011; Ares et al. 2010). However, the response of the understory community structure to thinning was dynamic (Table 8). New species appeared within a short period (1 to 2 years) after thinning but the number of new species appeared to decrease over time. For example, after thinning, 31 and 37 species appeared after 1 and 2 years post-thinning, respectively, while just 4 and 3 species appeared during the period from 3 and 7 years after thinning, respectively. In our study, light-demanding species such as S. faberi, C. formosana, and V. carlesii started to appear and become dominant in the shrub understory layer within 7 years following thinning. These results were consistent with those of Ishii et al. (2008) who reported that there was greater species richness of understory vegetation in thinned stands than in unthinned stands in C. japonica forests 4 and 6 years after thinning. However, He and Barclay (2010) found no significant effects due to thinning on the majority of understory species after 27 years in a 51-year-old Douglas fir plantation. Lei et al. (2007) also reported that understory diversity of larch-spruce-fir forests was similar between thinned and unthinned stands 12 years after thinning. Based on these results, we predict that thinning would have less of an effect on understory vegetation diversity in Chinese fir plantations with increasing amounts of time post-thinning. The immediate effects of thinning, which gave way to a slower succession of species ranging from several to dozens of years after thinning, were probably attributed to variation of the maturation process and rotation patterns following disturbance. Specifically, thinning clearly decreased canopy density and increased light transmittance to the understory, which is beneficial to the growth of some light-demanding species (Cañellas et al. 2004).

Increased diversity and cover of understory species may change microclimatic features, but the relationship between each other was unresolved. Thinning appeared to promote the shrub layer more than the herb layer, consistent with many previous studies that thinning expedites recovery of the understory shrub layer in some conifer forests (Chan et al. 2006; Dodson et al. 2008; Ares et al. 2010). We also found that soil capillary porosity, soil total porosity, and increased water content tended to be higher in thinned stands, allowing for higher water content in the soil (Table 9), as has been reported elsewhere (Gray et al. 2002; Griffiths et al. 2010). Increased water content may be attributed to decreased evapotranspiration and increased precipitation through gap openings after thinning. Understory development may have the potential to loosen soil through penetration and extension of roots, which may result in decreased soil bulk density in thinned stands. Griffiths et al. (2010) reported that the soil bulk density was greatly decreased in thinned stands, which is consistent with our findings. Conversely, the distribution of understory species has been directly linked to soil water content in Oregon’s Western Cascades (Lookingbill et al. 2004) and British Columbia (Wang 2000). Ares et al. (2010) reported that understory species could affect soil water status 11 years after thinning which suggested that understory species benefited from augmented water availability due to stand density reductions. However, full appreciation of the effects of thinning management practices on understory diversity requires measurement of key microclimatic factors within the stands, i.e., crown opening, light transmittance, soil physical and chemical properties, and understanding of the correlation between them.

The relationship between tree growth, soil status, and understory

The positive correlation between understory biomass and volume (individual volume and stand volume) indicated that the growth of both trees and understory vegetation (especially in the herbaceous layer) would respond positively to thinning management (Fig. 3). The productivity decline in successive rotations of Chinese fir in monocultures has become a critical productivity issue and needs to be resolved. Many people have attributed the decline in productivity more to the silvicultural practices employed (e.g., short rotation, whole tree harvest, poor management of understory vegetation) than the Chinese fir species itself (Ma et al. 2007; Tian et al. 2011a; Bi et al. 2007; Wu and Ma 2011). Chinese fir plantations are generally located on slopes of 20 % or more and are characterized by low mountains and hilly terrain with broken topography and complicated geology (Bi et al. 2007). Understory vegetation decreases and may disappear altogether if Chinese fir stands are allowed to remain closed for 3–5 years after planting. Therefore, soil water and nutrients are easily lost on steep slopes in the absence of understory, especially during the monsoon season (Yang et al. 1993). Thinning quickly restored understory vegetation diversity and biomass during the 1–2 years after thinning (Tables 6, 7 and Fig. 4), enhancing the soil’s water-holding capacity and nutrient content (Tables 9 and 10). Our results are consistent with other reports that the rapid increase of understory vegetation following thinning stimulates soil capillary development and decreased soil bulk density resulting from the penetration and loosening process of understory roots (Griffiths et al. 2010; Zhang et al. 2001). Sheng et al. (2005) reported that Chinese fir grows well when the soil bulk density is under 1.10 g cm−3, and does not grow well when the soil bulk density is over 1.30 g cm−3. The large understory biomass, especially within the herbaceous layer, produces large amounts of plant residue due to rapid nutrient turnover and accumulation of soil organic matter and nutrients (Selig et al. 2008).

Inquiry into whether fast-growing monoculture plantations are detrimental or beneficial for environmental conservation is not a trivial question. Attention should be given to the causes of declining forest productivity and environmental degradation. Sustainable forestry is based on forest management practices that promote optimal productivity while preserving the ecological balance required for long-term nutrient cycling and ecosystem stability (Embaye et al. 2005). Our study illustrated the reserved trees not only benefit from the augmented water and nutrient after thinning but also from increased understory diversity. The productivity and ecological balance of Chinese fir plantations were ultimately improved by proper thinning management due to the quick response of understory vegetation to thinning management. Maintaining a mature understory community is essential to maintaining ecological balance in Chinese fir stands in the long run. Therefore, thinning management at the young and middle stages is an effective method to enhance the forest environment and ease productivity declines in Chinese fir plantations. However, further studies should be carried out to evaluate the positive effects of understory vegetation after thinning on soil ecophysiological properties and forest microenvironment in order to establish sustainable management in these plantations in China.

Management implications

Our study has shown that thinning is an effective silvicultural management tool to foster a more structurally complex forest ecosystem over short time frames. For short rotation species (e.g., 20–25 years for Chinese fir plantations), the first thinning after the canopy closes (7–9 years) significantly increases the diversity and biomass of understory vegetation within a short time (1 or 2 years). Thus, understory improvement should have long-term implications for the enhancement of soil physical and chemical properties, because of the fast nutrient turnover and high nutrient concentration of the understory flora. In our study, the positive relationship between understory biomass and individual tree volume increment implies that growth of the understory after thinning should not decrease the stand volume increment. However, the understory vegetation and tree growth response to the second thinning (middle stage) was less pronounced than for the first thinning (young stage). Therefore, thinning should enhance understory vegetation development and benefit long-term soil productivity without competing with tree growth. Long-term investigations on the forest environment, forest yield, and their interaction after thinning need to be carried out in Chinese fir plantations in China.

Conclusions

Future management strategies in Chinese fir plantations should focus on ecointegrity and ecodiversity, rather than simply improving the stem volumes for the short term. The beneficial effects on understory composition and diversity resulting from thinning management were measurable 1–7 years post-thinning, especially within 1–2 years post-thinning. The soil water-holding capacity and nutrient availability were also enhanced by thinning within our study period. The positive effects of thinning on understory and soil were more profound in the young stage than in the middle stage. Thinning greatly increased QMD and individual volume, and the individual volume increment can be expected to compensate for the yield reduction resulting from any trees removed. Further research will characterize the effects of thinning on the interactive responses of understory development, soil, and growth of Chinese fir for long-term plantation management practices.