Jihua Hou · Junxia Tian · Li Xu · Zihao Zhang ·Zhi Chen · Nianpeng He

Abstract Evidence-based selective cutting at prescribed intervals as part of good forest management can enhance the carbon sequestration capacity of the forest. The Effect of forest management on carbon sequestration has, however,not been quantif ied. Thus, carbon content of various organs was measured for 323 tree species, 247 shrub species, and 233 herb species in seven temperate coniferous and broadleaved mixed forests that were subjected to selective cutting with restoration durations of 100, 55, 45, 36, 25, 14, and 6 years to explore dynamic changes in carbon storage. The results showed that biomass carbon allocation in diff erent organs followed a pattern: trunk > root > branch > leaf for all forests. With longer restoration durations, more carbon accumulated in diff erent organs and in soils. Interestingly,when the restoration duration exceeded 50 years, carbon storage in ecosystem was larger than that in primary forests with 100-year cutting intervals, suggesting that a reasonable selective cutting interval can increase forest carbon sequestration. Mean diameter at breast height (DBH) and forest carbon storage were signif icantly positively correlated, and carbon storage of selectively cut forests exceeded that of primary forests when the stand mean DBH exceeded 15.66 cm. Therefore, mean DBH of forests can be an indicator for combining sustainable forest management and forest carbon sequestration. Additionally, the classic coeffi cients of 0.45 and 0.50 used to estimate carbon sequestration underestimated values by 2.65% and overestimated by 8.16%,respectively, in comparison with the measured carbon content from diff erent plant organs.

Keywords Selective cutting · Forests · Diameter at breast height · Carbon · Storage · Management · Restoration

Introduction

Forests store a large amount of soil organic carbon (SOC)and are considered to be the most important carbon (C) sink in terrestrial ecosystems (Fang et al. 2001, 2018; Xu et al.2019). In recent decades, six major national ecological restoration projects have been implemented in China, yielding a signif icant increase in C sequestration equivalent to 50-70% of the total annual sink from all major terrestrial ecosystems (Piao et al. 2009; Lu et al. 2018; Xu et al. 2019).However, China still faces considerable pressure to reduce carbon emissions, with a goal of increasing forest stocks by 4.5 × 10 9 m 3 by 2030 compared to the 2005 levels. Forest C storage can be enhanced using sustainable management practices. Selective cutting, def ined as regularly and repeatedly harvesting trees in a specif ic area, is a key harvesting method for the sustainable management of natural forests in China (Xua et al. 2016). The appropriate intensity and intervals of selective cutting are estimated to promote the growth and regeneration of forests and to enhance their C sequestration. However, specif ic data is lacking regarding how forest management inf luences C sequestration.

Forest C pools mainly comprise live vegetative biomass,debris, and SOC (Tang et al. 2018; Xu et al. 2018). Selective cutting strategies have various dynamic inf luences on these C pools (Fig. 1). Cutting removes some live biomass,resulting in a reduction in aboveground tree biomass and a decrease in primary productivity in the early phase of restoration. However, selective cutting with an appropriate intensity can provide favorble light, space, and nutrient release for the residual tree species, and strongly promote their primary productivity (Vargas et al. 2009). Primary productivity tends to reach a peak when forests develop to middle age and then declines. Selective cutting also initially increases soil heterotrophic respiration following harvesting, followed by a decline through the restoration period. Total ecosystem respiration is highest when forests are relatively young,and declines as forests mature (Pregitzer and Euskirchen 2004). Selective cutting decreases vegetation C storage and increases the amount of woody debris and litter on the forest f loor. The decomposition of this debris results in declining detrital C storage, with this process at its highest rate immediately following selective cutting. The loss of detritus from the forest f loor is thought to proceed more rapidly than the accumulation of live biomass, resulting in an initial loss of total carbon storage. With longer intervals between selective cutting, plant growth and accumulated biomass increase, and detrital pools build up, resulting in an extended period of carbon accumulation (Piene and Vancleve 1978). Finally,the buildup of C in live biomass and C on the forest f loor results in mortality and decomposition losses, which moves the pools toward a steady state with net ecosystem productivity approaching zero (Carey et al. 2001).

Few studies have estimated the C storage of forest vegetation based on directly measured C content of diff erent plant organs and species due to the expense and labor-intensity of such an undertaking. Most studies have used the classic coeffi cients of 0.45 or 0.50 to estimate vegetation C storage,although some associated uncertainty is unavoidable (Piao et al. 2009; He et al. 2017; Lu et al. 2018; Tang et al. 2018).However, a recent study demonstrated that the C contents of leaves, branches, and roots of trees diff er greatly (Zhao et al. 2018), indicating there is broad uncertainty in estimating large-scale vegetation C storage with unknown margins of error. Diameter at breast height (DBH) is an important indicator of tree growth and is easy to measure in the f ield. If mean DBH accurately ref lects forest C storage in restoration following selective cutting, then this simple measure could serve as a proxy for more complex measurements and allow for the development of evidence-based practices for selective cutting in forest management combined with C sequestration and economic benef its.

In the present f ield investigation of seven temperate coniferous and broad-leaved mixed forests that were subjected to selective cutting with diff erent restoration durations (>100,55, 45, 36, 25, 14, and 6 years), we measured the C content of leaves, branches, stems, and roots of 803 plant species to explore dynamic changes in forest C storage. We aimed to (1) describe how C storage and distribution are altered with restoration duration after selective cutting in temperate coniferous and broad-leaved mixed forests, (2) demonstrate whether rational intervals of selective harvesting can enhance both C sequestration in the temperate forests of northern China and economic benef its, (3) quantify the error between the measured vegetation C storage and the estimates of vegetation C storage associated with the classic coeffi cients of 0.45 and 0.50, and (4) investigate whether stand mean DBH can be used as an indicator of forest C sequestration for future forest management that balances ecological and economic benef its.

Fig. 1 Theoretical framework describing the dynamics of forest carbon (C) storage after selective cutting to best manage forest C sinks.

Materials and methods

Study sites

The study was carried out in typical temperate coniferous and broad-leaved mixed forests in Jiaohe, Jilin Province in northeastern China (43°57′ N, 127°44′ E). The average monthly temperature is 3.8 °C with a maximum of 21.7 °C in July and a minimum of -18.6 °C in January, and average monthly precipitation of 695.9 mm. The soil is brown forest soil, supporting a rich diversity in plant species and complex community structure (Zhang et al. 2017). The main coniferous tree species isPinus koraiensisSieb. et Zucc.and the dominant deciduous broad-leaved tree species areJuglans mandshuricaMaxim.,Betula platyphyllaSuk. andAcer monoMaxim. The dominant shrub species areCorylus mandshuricaMaxim. andRhamnus schneideriLévl. et Vant.The dominant herbaceous plants areBrachybotrys paridiformisMaxim. ex Oliv. andVitis amurensisRupr.

Field sampling

Collection of plant and soil samples

We selected six forests with similar terrain and topography based on their selective cutting histories, with restoration durations of 55, 45, 36, 25, 14, and 6 years after previous selective harvesting in 1962, 1972, 1981, 1994, 2003, and 2011, respectively, and a forest with a restoration duration of at least 100 years, which we considered as a primary forest with a representative community structure. The harvesting intensity of these selectively cut forests was about 13 m 3 ha -1 . In each forest, we set up four 30 m × 40 m plots(Table 1). The DBH and height of all trees (DBH ≥ 1 cm)were measured and recorded. Two 5 m × 5 m quadrats and two 1 × 1 m quadrats were set up in each forest plot to measure the height and basal diameter of shrubs and the height of herbaceous species (He et al. 2018; Liu et al. 2018).

In each plot, three healthy individuals of each tree, shrub,and grass species were sampled. Sun-exposed and fully expanded leaves and branches with a diameter <1 cm were collected randomly; trunk material was gathered using an increment borer at breast height. Fine roots (<2 mm diameter) were also collected by removing the soil surrounding main roots until f ine roots could be distinguished from different species (Zhang et al. 2018a). In total, 323 tree species,247 shrub species, and 233 herb species were sampled in this manner. Simultaneously, the litter in every 1 × 1 m quadrat was collected. Soil samples were collected using a soil sampler in the 0-10 cm and 10-30 cm layers.

Determination of biomass

The DBH and tree height were used to calculate the biomass of trees, and basal diameter and height were used to calculate the biomass of shrubs according to allometric equationsdeveloped using f ield data from the same climatic zone and vegetation type (Wang et al. 2015; Zhang et al. 2018a). For herbs, all plants in the quadrats were harvested to obtain their biomass. Fresh samples were oven-dried at 60 °C for 48 h and weighed.

Measurement of C content of diff erent plant organs

Samples of leaf, branch, trunk, f ine root, and litter were carefully cleaned and oven-dried at 60 °C for 48 h. Soil samples were sieved using 2 mm mesh and air-dried, and visible roots and organic debris were removed by hand. All plant and soil samples were ground to a f ine powder using a ball mill (MM400, Retsch, Germany) and an agate mortar grinder (RM200, Retsch, Haan, Germany). The C content of all samples were measured using an elemental analyzer(Vario MAX CN Elemental Analyzer, Elementar, Germany)(Zhang et al. 2018a).

Calculation of C storage in vegetation, soil, and ecosystem

C storage in leaves, branches, trunks, and roots for each species was calculated as the biomass multiplied by the C content for each organ of a specif ic species. The C storage was summed for all parts and all plant species to obtain vegetation carbon storage (VCS, kg C m -2 ). The formula is as follows:

where,VCS is the sum of C storage of trees, shrubs, and herbs.mis the number of plant organs,i= 4 for trees (leaf,branch, trunk, and root),i= 3 for shrubs (leaf, branch, and root),i= 2 for herbs (leaf and root).B iandC iare the biomass and C content of plant organi. We also calculated C storage using the classic coeffi cients 0.45 and 0.50 to compare their inf luence on the errors of estimates.

Soil C storage was calculated according to the following formula:

where,SOCSis soil C storage per unit area (kg C m -2 ),S iis soil organic C content (%),M iis soil bulk density (g cm -3 ),D iis soil thickness (cm),G iis the percentage of the volume of gravel with a diameter greater than 2 mm (%), andkis the soil layer number. Soil bulk density was calculated using the classic pedotransfer function (Xu et al. 2015).

Ecosystem C storage was calculated according to the following formula:

where,ECSis the ecosystem C storage per unit area(kg C m -2 ),VCSandSOCSare the C storage in the vegetation and soil, respectively. In this study, with regard to theSOCS,we calculated the C storage in the soil at the 0-30 cm depth.

Data analyses

One-way analysis of variance (ANOVA) and Fisher’s least signif icant diff erence (LSD) multiple comparison were used to analyze diff erences in vegetation C storage, soil C storage,and ecosystem C storage among diff erent forests with diff erent restoration durations. A pairedttest was used to evaluate the diff erences in estimated vegetation C storage among the measured C content of diff erent plant organs as well as the 0.45 and 0.50 coeffi cients. Best-f it equations describing the relationship between restoration duration and vegetation,soil, and ecosystem C storage and equations describing the relationship between DBH and C storage were explored. Differences were regarded as signif icant atP≤ 0.05. All statistical analyses were performed using SPSS 17 (SPSS, Chicago,IL, USA) and plotted using SigmaPlot 12 (SystatSoftware,San Jose, CA, USA).

Results

Changes in C storage with restoration duration after selective cutting

For vegetation, C storage in trees, shrubs, and herbs were signif icantly diff erent among all measured sites. Storage was greatest in trees, followed by shrubs and herbs (P< 0.05).The C storage in trees with increasing restoration duration was ordered as 55 > 100 > 45 > 36 > 25 > 14 > 6 years. As expected, when the restoration duration exceeded 50 years,the C storage of trees was greater than that of primary forests(Table 2). C storage in diff erent tree organs in decreasing order was trunk > root > branch > leaf (Online Appendix Table S1).

For soil, C storage was not signif icantly diff erent among the forests with diff erent restoration durations (Fig. 2; Online Appendix Table S1). For ecosystems, however, C storage was signif icantly diff erent among all measured sites. Ecosystem C storage, similar to vegetation C storage, was signif icantly positively correlated with restoration duration(Fig. 2). Soil and trees stored the bulk of C pools, with soilsproviding 44.25%-66.51% of total C storage and trees storing 32.74-55.38% (Table 2; Online Appendix Table S1).

Table 2 Mean C storage in various types of vegetation according to restoration age

Fig. 2 Carbon (C) storage at ( a) vegetation, ( b) soil (0-30 cm), and( c) ecosystem levels in temperate forests in northern China based on restoration duration. C at 30 cm soil depth was used to estimate ecosystem C storage. Blue lines represent C storage in the primary forests at the various levels.

Relationship between stand mean DBH and forest C storage

There were signif icant correlations between ecosystem and vegetation C storage and stand mean DBH. Within a certain range, a greater mean DBH was associated with greater C sequestration (Fig. 3). When mean DBH was greater than 15.66 cm, C storage in selectively harvested forests after long-term restoration was greater than in primary forests(Fig. 3). However, mean DBH and soil C storage were not signif icantly correlated (Fig. 3 b).

Estimated errors from the diff erences in C content in plant organs

The results of the pairedttest showed that there were signif icant diff erences between the measured and predicted vegetation C storage based on diff erences in C content in diff erent plant organs (Table 3). Vegetation C storage calculated using the 0.45 coeffi cient was underestimated by 2.65%, compared with measured C content, and was overestimated by 8.16%using the 0.50 coeffi cient (Table 4). Estimated errors in estimated C storage varied with diff erent plant organs (Online Appendix Table S2).

Discussion

Inf luence of diff erent C content coeffi cients on the estimates of forest carbon storage

Fig. 3 Relationships between mean carbon (C) storage and stand mean diameter at breast height (DBH) at ( a) vegetation, ( b) soil(0-30 cm), and ( c) ecosystem levels in temperate forests in northern China. C at 30 cm soil depth was used to estimate ecosystem C storage. Blue lines in a and c represent stand mean DBH = 15.66 cm,which was calculated based on the f itted equation y = 2.091E0.093 x in a and the vegetation C storage of the primary forest

Our results indicate that the classic coeffi cient of 0.45 may underestimate forest vegetation C storage by 2.65% and 0.50 may overestimate storage by 8.16% in the cold-temperate forests in northern China. In most studies, vegetation C storage in forests is obtained from biomass by directly multiplying the C content by either 0.45 or 0.50. This method is simple and easy to operate at regional or global scales, but it is also an important source of uncertainty. The C content among leaves, branches, trunks, and roots of vegetation differs measurably, and even the C content of various plant organs among diff erent species and regions can be quite different (Zhao et al. 2018). Although it is diffi cult to measure the C content of various plant organs, such measurements should be encouraged for all forest ecosystems using existing f ield stations or through integrative data analyses. A system of C content coeffi cients for diff erent plant organs (Zhao et al. 2018) according to the type of forest will assist in developing more accurate estimates of forest C sequestration at diff erent scales from local to global.

Prescribed cutting intervals can enhance forest C sequestration

Selective cutting may inf luence the microenvironmental characteristics, stand growth, biomass allocation, and other outcomes and can have important impacts on forest C stoage. In this study, vegetation C storage gradually increased with restoration duration and was higher than that in primary forests with restoration durations greater than 50 years. After selective cutting, vegetation C storage in forests signif icantly decreased (Fig. 1) due to the direct reduction of tree density and biomass. However, light availability and individual tree nutrition improved and competition among trees decreased to some extent, allowing smaller trees to grow faster, resulting in rapid increases in the rate of productivity and biomass accumulation. Wang et al. ( 2013) found that C storage inLarix gmeliniiRupr. forests with a 26-year restoration duration was signif icantly higher than in primary forests in northeastern China. The change in vegetation C storage after selective cutting might be attributed to changes in tree species composition, stand age, climatic conditions, and restoration duration (Cheng et al. 2013). The practice of selective cutting has limited inf luence on shrub and herb layers,although C storage in shrubs and herbs accounted for only 0.50-2.22% of the total C storage in forests in this study. A past investigation demonstrated that 33 years after selective cutting, C storage in understory vegetation in Norway spruceforests was not signif icantly diff erent from that of primary forests (Nilsen and Strand 2008). The insignif icant change in C storage in the shrub and herb layers may result from forest growth, increasing canopy density, and the proximity of diff erent microclimates under the canopies.

Table 3 Effect of C contents on vegetation C storage estimation (the measured C content of various plant organs, 0.45, and 0.50, respectively)

Table 4 Estimated errors of vegetation C storage based on diff erent conditions of C content(measured C content of diff erent plant organs, 0.45, and 0.50,respectively)

In this study, we found that selective cutting intervals had no signif icant Effect on soil C storage in cold-temperate forests, which is supported by previous studies (Post et al.1982; Dixon 1994). However, there is no consensus on the impact of selective cutting on soil C storage. Some studies found that selective cutting reduced forest soil C storage(Piene and Vancleve 1978), while others found that selective cutting increased it (Vargas et al. 2009; Zhang et al. 2017).Factors such as climate, cutting intensity, vegetation composition and restoration time all aff ect the dynamics of soil C storage in forests, and the impact processes are complex (Su et al. 2006; Zhang et al. 2018b). Therefore, further research is needed on the impact of deforestation on soil C storage.

Vegetation and soil C storage are important indicators for evaluating the Effects of forest C sequestration (Levine et al. 1995). In this study, C storage in ecosystems was lower than in primary forests after short selective cutting cycles. During restoration, C storage of selectively cut forest ecosystems increased gradually, and even exceeded that of primary forests after 50 years. Studies have shown that,in the short term, the biomass C pool may be reduced and C storage of the ecosystem will be lower than that of the undisturbed forests (Wen and He 2016). However, C storage in the ecosystem can be increased to a level comparable to that of primary forests under long-term restoration(Pyorala et al. 2012). Some researchers have suggested that high timber production and high C storage can be simultaneously maintained in forest ecosystems and have proposed adopting evidence-based management practices that ref lect the natural development of the forest to support this (Thornley and Cannell 2000). Selective cutting,as part of this management, can be applied to increase timber production, biomass energy, and forest C storage simultaneously and can increase C balance and natural regeneration of high-density forests (Alam et al. 2013).Carefully developed selective cutting strategies can promote the growth of the remaining timber, regeneration of woody plants, and development of the stand structure;thus, the amount of C f ixation in the stand will be higher than that of nonselectively cut forests (Dwyer et al. 2010).In summary, selective harvesting at intervals allowing for more than 50 years of restoration can enhance carbon sequestration as well as confer economic benef its in the temperate forests studied here, although it should be noted that these intervals may vary among diff erent forest types,such as tropical or subtropical forests.

Stand mean DBH can be used as an indicator in sustainable forest management

Our results showed that there was signif icant positive correlation between stand mean DBH and vegetation C storage, however, the DBH showed no correlation with soil C storage, consistent with the positive correlation among DBH, tree height, and biomass found by others (Zhang et al. 2014, 2018b). Within a certain range, as DBH increases, biomass and tree height also increase gradually, resulting in a change in the space and light conditions in the forest. Trees with larger DBH can obtain more nutrients and have greater productivity and C sequestration compared to trees with smaller DBH. When the ratio of a forest’s C storage to that of primary forests is greater than 1, stand mean DBH can be regarded as optimized DBH(Thornley and Cannell 2000; Alam et al. 2013). This relationship could be used as a criterion to choose trees for harvest, increase timber production and ensure C storage in the stand. Moreover, mean DBH is easier to measure than many other variables, making it a convenient option for combining forest management with C sequestration.

Conclusions

The restoration duration of the selectively harvested forests was signif icantly correlated with the C storage of the forest vegetation and ecosystem but was not signif icantly related to soil C storage. Therefore, rational intervals of selective harvesting with restoration periods of more than 50 years could both enhance C sequestration in temperate forests of northern China and support economic interests. However, the quantitative def inition of reasonable cutting intervals may diff er in other forest types, such as tropical forests or subtropical forests. In cold-temperate forests, stand mean DBH can serve as a straightforward means of informing selective harvest strategies, and the relationship between mean DBH and C storage can help to coordinate practices for optimal forest management and C sequestration. Furthermore, our f indings demonstrate that the classic coeffi cients of 0.45 and 0.50 may, respectively, underestimate forest vegetation C storage by 2.65%and overestimate it by 8.16% in these temperate forests,indicating that a system of C content coeffi cients for various plant organs in typical forests is necessary to decrease such estimation errors at local and global scales. Finally,our f indings provide new evidence to inform selective cutting practices that benef it both ecological and economic outcomes.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affi liations.