Effects of Thinning on Carbon Storage in a Mixed Broadleaved Plantation in a Subtropical Area of China
1. Introduction
Forest ecosystems provide multiple services, especially their ability to capture and store carbon (C). Current studies indicate that a forest ecosystem constitutes a net C sink that offsets ~25% of yearly anthropogenic C emissions, thus mitigating climate change actively [1,2]. In China, the area of plantations has expanded to 8.0 × 107 ha in 2021, which translates into an increased terrestrial carbon sink in biomass and soils [3]. Due to poor ecological stability and a decrease in carbon sequestration capacity of monocultures [4] and the higher biodiversity, resistance, and resilience to disturbances of mixed forests [5,6,7], mixed plantations have been proposed and implemented since the last decade. Even though forestation sequesters carbon, the scalability of this land use for meeting warming limitation targets has been questioned due to the sheer amount of land area required [8]. Furthermore, forestation of arable land results into trade-offs with local food production [9,10]. Therefore, since the Paris climate summit in 2015, forest management has been promoted as an efficient option to increase the accumulation of C by forest ecosystems [11]. Thinning is an essential silvicultural practice widely used in forest management [12], and it has been proved to promote the soil carbon storge of forest ecosystems by altering substrate charateristics [13,14,15,16,17] and the carbon metabolism process rate [18]. For instance, thinning induces more litterfall input, which initially alters substrate availability for microbial, soil enzymatic activity, and carbon mineralization [18]. Thinning through removing trees to open the canopy, which causes more solar radiation to reach the forest floor, further stimulates understory vegetation diversity [19], the litter decomposition rate, and soil microbial activities [20], which further change the carbon metabolism process [21,22,23]. Studies have shown different effects of thinning on the carbon storage of tree layers. Thinning reduces the carbon storage of a tree layer by removing trees, even though the promoted growth of the preserved trees cannot compensate for the removed biomass carbon in a short time [24,25]. Studies have also shown that thinning enhances the tree layer’s carbon storage in plantations but reduces the carbon storage of the litter layer and soil [26].
So far, studies of the effects of thinning on tree growth, understory diverstiy, soil carbon storage, and forest biomass carbon have mainly involved pure forests. In addition, thinning effects on forest carbon storage may be reinforced, counteracted, or offset by the thinning-induced soil respiration rate, removed tree biomass carbon, or a disturbed soil enviroment. Thus, it is uncertaion how different thinning intensities affect the carbon structure post-thinning, and much less is known on the effect of thinning, especially the different intensities of thinning on carbon storage dynamics in mixed broadleaved plantations. We assessed the carbon dynamics in an 11-year-old mixed broadleaved plantation of Schima superba Gardn. et Champ. × Castanopsis hystrix Miq. × Michelia macclurei Dandy after different intensities of thinning. The main objective of this study was to find a proper thinning intensity to enhance the carbon storage of the mixed plantation. Our specific aims were (1) to evaluate the effects of different intensities of thinning on the carbon structures in the mixed plantation; (2) to estimate the influnces of thinning on species-specific carbon growth; and (3) to describe the effects of varing intensities of thinning on the overall carbon storage in the mixed plantation.
2. Materials and Methods
2.1. Study Site and Experimental Design
The study site is located in Zhaoqing city in Guangdong, China (E 111°22′~E 112°06′, N 23°07′~N 23°25′). It belongs to a subtropical monsoon climate, with a mean annual precipitation of 1428.5 mm~1638.3 mm and a mean annual temperature of 21.5 °C (Figure 1).
The planted density in 2010 was 856 trees·ha−1, with 237 trees·ha−1 Castanopsis hystrix Miq., 548 trees·ha−1 Schima superba Gardn. et Champ., and 71 trees·ha−1 Michelia macclurei Dandy. They were randomly mixed and planted regularly in the field. In the first three years after planting, understory plants were removed manually to improve the survival rate of seedlings. In 2013, silvicultural management had ceased, and the three tree species started to sprout several seedlings. In 2021, the density was 1585 trees·ha−1, with 368 trees·ha−1 Castanopsis hystrix Miq., 1101 trees·ha−1 Schima superba Gardn. et Champ., and 116 trees·ha−1 Michelia macclurei Dandy. The soil type is Latosol, and the forest floor is a typical moder.
Twelve square plots, 625 m2 each, were established in July 2021. We thinned the nine plots with 3 intensities, namely, 20%~30%, 31%~40%, and 41%~50%, in August 2021, and three other plots were untreated (control). There were three replicates for each thinning intensity. We adopted the method of thinning from below. The thinning intensity was the percentage of the basal area of the harvested trees in the total basal area of the original stands. Characteristics for each plot are presented in Table 1. A standard protocol for tree data collection (diameter, heights of trees, and crown length) was applied.
2.2. Understory Vegetation Survey and Index Calculations
In order to investigate the diversity of the understrory plants, in July 2021 and August 2022, 5 subplots (2 m × 2 m per subplot) were set up, with the center and four corners of each plot located. We recorded the shrub and herbaceous species as well as the quantity of each species. The coverage (in % of the plot area) of all understory plant species was recorded. For the shrub species, numbers, ground diameter (2 cm above the soil surface) and height (vertical height) were measured in the field. For herbaceous plants, the height was only measured for species with cover values > 2%. Species diversity was estimated by the Simpson index (Equation (1)), species richness was calculated by the Margalef index (Equation (2)), the species diveristy change index was estimated by the Shannon–Wiener index (H′) (Equation (3)), and the species evenness index was calculated by the Pielou index (Equation (4)).
where Ds is the Simpson diversity, pi is the proportion of species i in the plot, and S is the total species of shrub or herbaceous plants; DM is the Margalef richness index; N′ represents the number of individuals within the sample plot; H′ is the Shannon–Wiener index; and E is Pielou’s evenness index.
2.3. Litter, Soil Sampling, and Measuring Methods
Fresh litter was collected in July 2021 and August 2022 in a quadratic, square frame of 0.0625 m2. We divided the plot into four equal squares and sampled the litterfall at the centers of the squares. Once dried at 40 °C until constant, the different fractions (leaves, beech nuts, seed capsules, twigs, bud scales, etc.) were separated manually. Here, only the leaf fraction results are reported.
Soil samples were collected using a soil corer with a diameter of 8 cm in July 2021 and August 2022. Three replicates were taken for each treatment and soil depth (humus layer, 0~20 cm) in the plot. These samples were collected along four transects. The distance between sampling locations was 5 m in the plots.
After removing root particles, litterfall, the humus layer, and mineral soil were dried at 40 °C. Mineral soil was sieved (27].
2.4. Leaf Area INDEX Measurement
To measure the the leaf area index (LAI), four hemispherical canopy photographs were taken skyward from the forest floor with a fisheye lens at 4 locations, where the litterfall was collected in each plot. We applied Gap Light Analyzer Version 2.0 (GLA) software [28] to analyze the hemispherical canopy photographs to get the LAI.
2.5. Biomass Carbon, Litter Carbon, and Soil Carbon Calculations
The tree biomass was estimated for each plot using the species-specific allometric equations presented in Table 2. The aboveground biomass of Castanopsis hystrix was the sum of the biomass of the trunk, bark, branches, and leaves. The aboveground biomass of Michelia macclurei was the sum of the biomass of the trunk with bark, branches, and leaves. The carbon storage of trees was calculated from the biomass and carbon concentration.
Soil organic carbon storage was calculated as follows:
where Ci, SOCi, and BDi represent SOC stocks (t·ha−1), SOC concentration (g·kg−1), and soil bulk density (g·cm−3) in the i soil layer (cm), respectively. The BD values were calculated as [29]:
Carbon storage of litter was calculated from the dry weight and SOC conent.
Table 2.
Species-specific biomass allometric equations and C concentrations.
Table 2.
Species-specific biomass allometric equations and C concentrations.
| Species | Biomass (kg) | C Content Percentage | References | |
|---|---|---|---|---|
| Schima superba | Aboveground | W = 0.5373 × D1.903 (R2 = 0.91) | C = 0.5077 | [30] |
| Roots | W = 0.5759 × D1.4093 (R2 = 0.79) | C = 0.5487 | ||
| Castanopsis hystrix | Trunk | W = 0.0641 × (D2H)0.8699 (R2 = 0.99) | C = 0.4902 | [31,32] |
| Bark | W = 0.0105 × (D2H)0.8246 (R2 = 0.92) | C = 0.4491 | ||
| Branches | W = 0.0001 × (D2H)1.3949 (R2 = 0.81) | C = 0.4912 | ||
| Leaves | W = 0.0000028 × (D2H)1.6052 (R2 = 0.91) | C = 0.5018 | ||
| Roots | W = 0.1210 × (D2H)0.6495 (R2 = 0.81) | C = 0.4775 | ||
| Michelia macclurei | Trunk with bark | W = 0.033232 × (D2H)0.97166 (R2 = 0.98) | C = 0.5059 | [33,34] |
| Branches | W = 0.022721 × (D2H)0.84435 (R2 = 0.96) | C = 0.5050 | ||
| Leaves | W = 0.079679 × (D2H)0.59671 (R2 = 0.96) | C = 0.5077 | ||
| Roots | W = 0.039307 × (D2H)0.86499 (R2 = 0.96) | C = 0.4232 |
2.6. Data Analysis
Homoscedasticity was tested using the Fligner–Killeen test, and normal distribution was tested using the Shapiro–Wilk test. After logarithmic transformation, the data for the carbon storage of trees presented a normal distribution and homoscedasticity. To test for differences in the understory aboveground biomass between treatments, an ANOVA was used. To ascertain differences within the tree, soil, litter layer, and total carbon storage, respectively, Tukey–Kramer’s HSD test was applied. All statistical tests were performed with R, version 4.2.2 (QUOTE: R Development Core Team 2022).
4. Discussion
Forests sequester CO2 in the form of biomass and soil carbon. Forest C pools contain about 73% of the global vegetation carbon storage, with 44.5% of forest C stock in soil in the top meter [35]. In this study, the carbon structure was around 70% in soil, 20% in tree biomass (above- and belowground biomass), and 10% in the litter (Table 5). These values indicate that carbon was mainly stored in the tree layer and soil layer [36], accounting for over 90% of the total carbon storage in the plantation. There is a fundamental difference in their carbon structures due to different compositions of tree species. This is because forest communities–assemblages of tree species in a stand vary in their capacity to capture and store carbon [36]. The C storage in plantations of fast-growing tree species is significantly higher than that in plantations of slow-growing tree species. For example, under similar habitat conditions and management measures, the C storage in 27-year plantations of M. macclurei (359.43 t/hm2) and Mytilaria laosensis Lecomte (319.80 t/hm2) were significantly higher than in C. hystrix, Pinus massoniana Lamb., and Mesua ferrea L. plantations (225.87 t/hm2, 222.43 t/hm2, and 207.81 t/hm2) in subtropical China [37]. The 11-year-old plantation was composed of 60% S. superba, 30% C. hystrix, and 10% M. macclurei, with a higher tree biomass carbon growth for M. macclurei and S. superba and a lower carbon growth of C. hystrix (Figure 2D). One year after, the total carbon storage values, including tree layer biomass carbon, the litter layer carbon, and the soil layer carbon, were reduced in our study; however, the total carbon in the 20%~30% intensity thinned plots was less reduced in the control plots. Thinning by removing trees relieves the competition of trees, and then trees grow faster in conditions of low competition, especially fast-growing species [38]. The other possible explanations for the lower response of C. hystrix and S. superba to thinning may reside in the fact that there were more C. hystix and S. superba trees in the plots and that thinning led to a decrease in intraspecific competition for the three tree species, whereas interspecific competition was not reduced enough for the C. hystrix and S. superba trees [12]. Our study indicates that it is necessary to combine fast- and slow-growing species when planting mixed forests and to apply a moderate intensity of thinning to prevent the loss of carbon storage.
Forest management, such as tree pruning, thinning, the use of lime, fertilizer application, irrigation, and site preparation intensity, has received increasing attention due to its predictable effects on ecosystems, especially on the content of soil C. Changes in forest management practices have reportedly resulted in a significant loss of SOC over the past two centuries [39]. In our study, the total C storage in the mixed broadleaved plantation decreased, due to the losses of soil C (18.83 t·ha−1·yr−1) and litter C (2.53 t·ha−1·yr−1), even though the tree biomass C had a 3.60 t·ha−1·yr−1 increment. Due to the SOC containing more than three times the amount of organic carbon as in the overlaying vegetation [40], a slight change may affect the carbon sequestration of forest ecosystems. In 41%~50% thinned plots, the proportion of soil C (Table 5, 74.25%) decreased, even though the tree biomass C and litter C increased. That means that a higher intensity of thinning causes greater disturbance to forest soil. Forest soil is the medium in which forest plants evolve, grow, and derive their nutrients and water supply [12,41]. As a result, edaphic factors play a larger role in plant diversity [42]. In our study, the understory diversity and richness increased with the increase in thinning intensity (Figure 6), which were negatively related to the litter C (Figure 7). An increase in the species richness of plants leads to an increase in the number of available micro-niches and an increase in microbial diversity [43], further promoting the decomposition of C in the litter layer, humus layer, and soil layer. The competition for nutrients between plants and microorganisms facilitates microbial decomposition of litter in forest ecosystems [44] and more microbial residues. In the topsoil, the plant residues and microbial residues regulate the SOC storage [45].
There is an advantageous growth at low density, as commonly observed in selective or future crop tree thinning systems for individual tree size growth acceleration [46]. Competition reduction enables an increase in stand density, mass production, and climate change mitigation through higher carbon storage [47,48,49]. Thinning reduced the tree layer C (Figure 1A) by removing the trees but promoting the tree biomass C growth by relieving the competition among trees for resources (Table 6). The tree layer biomass carbon was significantly positively related to the leaf area index (LAI) and the tree layer coverage (Figure 7). In the 41%~50% thinned plots, the tree layer biomass C was higher than in the other thinned plots, due to the highest LAI (1.68 ± 0.3) and coverage (69.19 ± 4.5). The litter C was positively related to the coverage and tree layer biomass C, with the highest litter C in the 41%~50% thinned plots as well. That may be caused by the highest diversity and richness of understory plants [50], which is induced by the opening of the forest canopy and then causes more solar radiation to reach the forest floor, further stimulating understory vegetation diversity [18].
5. Conclusions
The results of this study indicated that to promote carbon sequestration, a mixed plantation should be composed of fast-growing and slow-growing tree species. The different intensities of thinning in a mixed broadleaved plantation improved the growth of reserved tree biomass carbon, promoted the diversity of understory vegetation, reduced the litter layer carbon, and reduced the total carbon in the short term. The 20%~30% thinning intensity promoted carbon sequestration, while the greater thinning intensity than 30% reduced the total carbon of the forest ecosystem. In our study, the promoting effect of thinning on growth was mainly caused by the change in the aboveground environment and soil nutrient availability, and the effect of thinning varied with the intensity of thinning. According to the research, during the reforestation process, the fast-growing and slow-growing tree species should be planted with a proper ratio to enhance carbon sequestration. In the long term, in order to promote forest ecosystem carbon storage, a higher intensity of thinning should be applied.
