Converting larch plantations to mixed stands: Effects of canopy treatment on the survival and growth of planted seedlings with contrasting shade tolerance

Forest Ecology and Management 409 (2018) 19–28

Contents lists available at ScienceDirect

Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco

Converting larch plantations to mixed stands: Effects of canopy treatment on the survival and growth of planted seedlings with contrasting shade tolerance

MARK

⁎

Deliang Lua,b,c, G. Geoff Wangd, , Jinxin Zhanga,b, Yunting Fanga,b, Chunyu Zhua,b,c, ⁎ Jiaojun Zhua,b, a

Qingyuan Forest CERN, Chinese Academy of Sciences, Shenyang 110016, China CAS Key Laboratory of Forest Ecology and Management, Institute of Applied Ecology, Shenyang 110016, China University of Chinese Academy of Sciences, Beijing 100049, China d Department of Forestry and Environmental Conservation, Clemson University, Clemson, SC 29634, United States b c

A R T I C L E I N F O

A B S T R A C T

Keywords: Plantation conversion Canopy retention Artificial regeneration Seedling survival Seedling growth Carbohydrate storage

Larch (Larix spp.) plantations are important for timber production in Northeast China, but this monoculture practice has led to problems such as decreased soil fertility and water-holding capacity. To examine the possibility of gradually converting pure larch plantations to mixed stands by small-scale canopy regulation, we planted seedlings of two species with contrasting shade tolerance, light-demanding Manchurian walnut (Juglans mandshurica Maxim.) and shade-tolerant Korean spruce (Picea koraiensis Nakai), in larch plantations with four different canopy retention intensities (larger gap, 160 m2; smaller gap, 45 m2; thinning, 25% intensity based on basal area; and control, forest understory). After two growing seasons, we found that both species had higher survival rates and growth rates in larger gaps than in forest understories, but the detailed responses to treatments differed between species. Manchurian walnut responded strongly to larger gaps but insensitively to other treatments, especially with respect to biomass accumulation. In contrast, Korean spruce responded gradually with increasing canopy openness. However, canopy treatments had almost no effect on non-structural carbohydrate (NSC) concentration, biomass allocation, and NSC pool allocation, which only differed between species. Our findings indicated that the two species of contrasting shade-tolerance were able to survive and grow in larch plantations, and a small-scale canopy treatment, especially creating gaps of ∼160 m2 in size, could significantly improve seedling survival and growth during the first two years. Therefore, enrichment planting in conjunction with a low-intensity canopy regulation may play an effective role in converting larch plantations to mixed stands while maintaining continuous stand functions during the conversion process.

1. Introduction Larches (Larix spp., mainly L. olgensis Henry and L. keampferi (Lamb.) Carr.) are important fast-growing commercial species in Northeast China, which have been widely planted for several decades to replace natural secondary stands to meet timber demands (Wang and Liu, 2001; Zhu et al., 2008). Currently, there are about 2.6 million ha of larch plantations in Northeast China (Liu et al., 2005; Yan et al., 2017). Although larch plantations play an essential role in timber supplies, such long-term monoculture practice leads to potential problems. Compared with the adjacent natural secondary stands, the larch plantations are facing reduced soil fertility. For example, after a conversion

from natural secondary stands to larch plantations, soil organic carbon and soil microbial biomass declined significantly (Chen and Yin, 1990; Chen and Li, 2003; Wang and Wang, 2007; Yang et al., 2010). An increased evapotranspiration in larch plantations also decreased the soil water content (Yang et al., 2010), which could potentially alter hydrological cycle. Moreover, larch plantations may be more vulnerable to disease, pest insects, or other disturbances because of the homogeneous canopy structure (Xu et al., 2000; Li, 2004; Haughian and Frego, 2016) and show less self-recovery capability (Roberts, 2004; Zhu et al., 2010). Thus, there is an urgent need for improving the resource sustainability and ecological functions of larch plantation in Northeast China (Mason and Zhu, 2014).

⁎ Corresponding authors at: Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang 110016, China (J. Zhu). Department of Forestry and Environmental Conservation, Clemson University, Clemson, SC 29634, United States (G.G. Wang). E-mail addresses: [email protected] (G.G. Wang), [email protected] (J. Zhu).

https://doi.org/10.1016/j.foreco.2017.10.058 Received 28 August 2017; Received in revised form 29 October 2017; Accepted 31 October 2017 0378-1127/ © 2017 Published by Elsevier B.V.

Forest Ecology and Management 409 (2018) 19–28

D. Lu et al.

opportunities to survive extreme conditions such as shade (Kobe, 1997; Myers and Kitajima, 2007). Seedlings with different functional types could have different carbohydrate allocation strategies, which may contribute to the coexistence of different species (Zhang et al., 2013; Villar-Salvador et al., 2015). Some studies also reported that seedlings had plasticity and could trade off in disadvantageous stand conditions. For example, Sevillano et al. (2016) evaluated the biomass allocation of European beech (Fagus sylvatica L.) under different light intensities and found that seedlings allocated more biomass to above-ground growth in shaded environments. However, other studies claimed that the biomass allocation of European beech was not affected by light availability (Curt et al., 2005) but mostly by ontogenic traits (Van Hees and Clerkx, 2003). The inconsistent results indicate that factors affecting seedling performance may differ among sites with different stand conditions, and intra- or inter-specific differences may also result in different outcomes. The objective of our study was to evaluate the possibility of converting pure larch plantations to mixed stands through enrichment planting under low-intensity canopy treatments. We selected two local tree species, Manchurian walnut (Juglans mandshurica Maxim.) and Korean spruce (Picea koraiensis Nakai). Manchurian walnut is a dominant broadleaved light-demanding species in Northeast China. It is also a commercial species and has important medicinal values. Korean spruce is a coniferous shade- and drought-tolerant species, which is distributed in most natural stands in Northeast China. We expected that Manchurian walnut and Korean spruce seedlings would coexist in the regeneration layer and gradually develop into different canopy layers due to their different growth characteristics. Our specific objectives were to (1) compare seedling survival and growth in response to different canopy treatments; (2) compare seedling strategy of carbohydrate storage allocation in response to different canopy treatments; and (3) find a minimal intensity but effective canopy treatment approach which could promote the regeneration of these two seedling species with contrasting shade tolerance in larch plantations.

As forest management strategies shift from pursuing timber production to providing ecological services (Li and Zhou, 2000), the approach of “close-to-nature” silviculture is becoming increasingly popular. As a result, thinning and group harvesting are commonly used to emulate fine-scaled natural disturbances and provide suitable microenvironments for tree regeneration (York et al., 2004; Zhu et al., 2010; Hu et al., 2012; Knapp et al., 2013). Previous studies have considered promoting natural regeneration in larch plantation understories (Deal, 2007; Yan et al., 2013). For example, a thinning experiment assessed the possibility of converting even-aged larch (L. olgensis) plantations to uneven-aged stands and found that natural L. olgensis seedlings failed to survive the current year after germination under different thinning intensities, but some broadleaved tree species could establish successfully (Zhu et al., 2010). These results indicate that thinning intensities may limit natural larch regeneration (Liu, 1997; Wang and Zhang, 1990; Zhu et al., 2008), but it may be possible to convert pure larch plantations to mixed larch-broadleaved stands (Zhu et al., 2010). However, the natural regeneration of broadleaved species could be poor after several years even with thinning treatments (Lei et al., 2003), which may be a result of increased competition from herbs and shrubs (Man et al., 2009; Kern et al., 2013). In addition, treatment responses were not only related to light levels (Thomas et al., 1999) but also other environmental factors, such as disturbance history and stand spatial pattern (Lei et al., 2003). For example, larch plantations located on the downslope of secondary stands could be more feasible for the establishment of some broadleaved species whose seeds are more easily moved downslope (Yan et al., 2013). Animals such as rodents could also affect seed dispersal and then influence seedling establishment (Wang et al., 2017). Given the uncertainty of natural regeneration, a combined treatment of canopy manipulation and seedling planting may be required to establish target species in larch plantations. Enrichment planting in conjunction with thinning or group harvesting may greatly shorten the time needed to convert pure plantations to mixed stands compared with natural regeneration (Paquette et al., 2006; Owari et al., 2015). Owari et al. (2015) monitored the height growth of planted Korean pine seedlings in larch stands with different strip-cut widths and found that seedling height increased in wider strips. However, a strong harvesting intensity or areas harvested may be critically restricted in some stands due to the requirement of continuously providing desired forest services, such as water resource conservation (Wang et al., 2013) and wildlife habitats (Knapp et al., 2013). Furthermore, competition from understory shrubs and herbs could increasingly restrain seedling growth as canopy removal intensities are increased (Paquette et al., 2006), which suggests that competition control measures may be needed during the early stage of seedling establishment (Kern et al., 2013). Growth response to canopy treatments also varied among different planting species (Kobe and Coates, 1997), which provides flexibilities of species selection to meet stand conditions and management objectives (Newsome et al., 2016). A long-term study assessed the effects of gap size on the survival and growth of three planted tree seedlings in mixed coniferous stands and found that both seedling survival and growth increased with gap size, but species-specific differences existed (Newsome et al., 2016). For example, the growth of Engelmann spruce (Picea engelmannii Parry ex. Engelm.) was less sensitive to increased gap size compared with lodgepole pine (Pinus contorta Dougl. Ex Loud.) and subalpine fir (Abies lasiocarpa (Hook.) Nutt.) (Newsome et al., 2016). Therefore, a combination of different planting species and canopy treatments may achieve a multiple-layer understory structure. Seedling performance in the regeneration layer largely decides the extent to which the species can be used for planting. In addition to simple growth indicators such as height and diameter, carbohydrate storage could provide more knowledge on seedling adaption to different canopy environments. Non-structural carbohydrate (NSC) storage reflects the balance between carbon supply and demand (Sala et al., 2012). Seedlings with higher NSC storage may have more

2. Methods 2.1. Study site Our study was carried out in Qingyuan Forest, one of the Chinese Ecosystem Research Network (CERN) sites established by the Chinese Academy of Sciences. The study area is located in a mountainous region of Liaoning Province, Northeast China (41°51′N, 124°54′E, 500–1100 m above sea level). It has a temperate continental monsoon climate, with a warm and humid summer and a dry and cold winter. The mean annual air temperature is between 3.9 °C and 5.4 °C, with the coldest month of January averaging −12.1 °C and the warmest month of July averaging 21.0 °C. The annual precipitation is 811 mm, 80% of which falls during the summer. The frost-free period is 130 days, with the first frost in early October and the last frost in late April. The soil is a typical brown soil (25.6% sand, 51.2% silt, and 23.2% clay) which belongs to Udalfs according to the US Department of Agriculture soil taxonomy (second edition) (Yang et al., 2013). The soil thickness ranges from 30 cm to 60 cm (Zhu et al., 2008). The mean soil pH of 0–5 cm in larch plantations is 5.4. The soil total carbon in 0–5 cm is 43.7 g kg−1. The soil organic matter in 0–5 cm includes 73% heavy fraction carbon, 13.7% light fraction carbon and 13.8% mineralizable carbon (Yang et al., 2013). More detailed soil characteristics are shown in Yang et al. (2013). Historically, the area was covered by mixed broadleaved-Korean pine stands. However, decades of unregulated timber exploitation severely disrupted the forest ecosystems, and more than 70% of the stands became secondary stands, with almost no remaining Korean pine trees. Since the 1950s, larch has been extensively planted in patches to replace secondary stands to meet the growing timber demands as it is a fast-growing commercial species. Currently, natural secondary stands 20

Forest Ecology and Management 409 (2018) 19–28

D. Lu et al.

columns. In each sub-plot, 12 seedlings were planted (i.e., six seedlings per column) for the two species. Seedlings were watered one time immediately after planting because of an abnormal seasonal drought event in the current year, although watering was not included in our original experimental design. About 300 ml water was used for each seedling. Planting began in late April 2015 and ended in early May 2015. All seedlings were bare-root seedlings and individuals of the same species were transplanted from the same local nursery which also provided local seed and seedling sources for the forest farms. More seedlings than needed were prepared during the planting, which assured us to eliminate seedlings far from the average conditions. During the two-year experiment period, seedlings within the sub-plots had enough space to grow. To promote seedling establishment and growth and reduce competition, natural regeneration such as herbs and shrubs in all subplots (including sub-plots in the forest understory) were removed monthly. We weeded in a relatively high frequency to assure the planted seedlings were not affected by vegetation competition during the first two growing seasons. The two planting species were selected based on the following reasons: (1) Manchurian walnut and Korean spruce are typical lightdemanding species and shade-tolerant species, respectively. The performance of other dominant tree species may be predicted according to the results of these two species. (2) Both species have a large distribution in Northeast China, North Korea and Russia. Thus, the findings of our study may be applicable over a large area. (3) Both species are important timber species, and Manchurian walnut also has medicinal values.

accompanying mosaic plantations form the typical temperate secondary forest ecosystems. The species composition mainly includes Acer spp., Betula platyphylla Suk., Fraxinus mandshurica Rupr., Manchurian walnut, Populus davidiana Dode, Quercus mongolica Fisch. ex Ledeb., Tilia spp., Ulmus spp., etc. (Zhu et al., 2007). The common plantation species are larches, Korean pine, and Korean spruce. 2.2. Experiment design Our study used a complete randomized design. We selected larch stands (∼25 years) with similar topography and south-facing slopes between 15° and 25°. Trees were between 18 and 20 meters high, and the mean DBH was 18 cm. Four canopy treatments were included: control (forest understory; 15 m × 15 m plot); thinning (single-tree selection with a 25% intensity based on basal area; 50 m × 15 m plot); smaller gap (multiple-tree selection; harvesting about four trees to create expanded gaps with an average size of 45 m2); and larger gap (group harvesting; harvesting about seventeen trees to create expanded gaps with an average size of 160 m2). Our treatment intensities were relatively low compared to many published studies because the local forestry administration agency restricts the cutting area (no more than 200 m2) before harvesting in larch plantations. The limited cutting area is within the lower gap size limits according to a recently published study (Zhu et al., 2015). The gap and control treatments had four replicates, while the thinning treatment had three replicates. A total station (TKS-202, China) was used during gap creation to delineate gap shapes, calculate gap sizes, and located the sub-plots for seedling planting. All trees and shrubs within gaps were cut and removed from the stands, but only thinned trees were removed in thinning plots. We used chainsaws instead of large machinery during harvesting because only a small number of trees were cut to create the canopy environment for the experiment. In this way, we could also minimize the potential negative effects of harvesting on soil layer and other trees, and save more time and money. All harvesting work was completed in March 2015. Following timber harvest, 3 m × 3 m sub-plots with a 0.2-meterwide buffer zone were set up for seedling planting. In each control and thinning plot, one sub-plot was set up in the center or near center location to keep the microenvironments homogeneous between replicates. In each smaller gap plot, five sub-plots were established, including one sub-plot located in the gap center representing canopy gap area. The other four sub-plots were located in gap edge locations in the north, south, east and west under the canopy of gap border trees, representing the expanded gap area. In each larger gap, additional four sub-plots were set up in the middle of the gap center sub-plot and each gap edge sub-plot, which were within the canopy opening and also represented canopy gap area. The number of sub-plots in each treatment varied according to the degree of environmental variation. For example, nine sub-plots along two cardinal directions were used within a larger gap to represent the average gap condition because the environment within gaps is heterogeneous. Reduced sub-plot number in smaller gaps was due to the limited size. All vegetation and litter (mainly larch needles) within the sub-plots were removed before planting. Two local seedling species, one-year-old Manchurian walnut and three-year-old Korean spruce, were planted. Species of different ages were selected according to the silvicultural practices of local forest farms. Older coniferous seedlings are widely used for reforestation because of their relatively low growth rates compared with broadleaved seedlings. The height, root collar diameter (RCD), total biomass, and NSC concentration of Manchurian walnut and Korean spruce were, respectively, 31.0 cm versus 20.3 cm, 6.3 mm versus 6.8 mm, 5.9 g versus 13.5 g, and 75.1 mg g−1 versus 83.2 mg g−1. Each sub-plot was evenly divided into four columns along a north-to-south direction, and each species was assigned to two columns according to a random number, which avoided the same species being planted in adjacent

2.3. Data collection We examined seedling survival at the end of each growing season in 2015 and 2016. A seedling was considered dead if no green tissue was found on the stem (Obrien et al., 2014). For each sub-plot, the seedling survival rate was estimated by the ratio of surviving seedling number and total seedling number (i.e., twelve) for each species. Seedling survival rates in 2016 may be underestimated because two seedlings were harvested in each sub-plot in 2015. We regarded these two seedlings as dead when calculating the survival rates to achieve a conservative result. Following survival census, we measured the root collar diameter (RCD) and height of all surviving seedlings in each sub-plot. The RCD was measured using a digital caliper, and seedling height was measured as the distance from the root collar to the tip of the top bud. The RCD and height data of each seedling species within the same sub-plot were averaged respectively to represent the growth condition. Following RCD and height measurements, two surviving seedlings of each species within each sub-plot were harvested and transported using cool boxes to the laboratory for dry biomass and NSC analysis. Three Manchurian walnut seedlings were harvested in four sub-plots in the second growing season because of a relatively large seedling variation within the sub-plot. In the laboratory, all collected seedlings were divided into leaves, stems, and roots, and dried at 70 °C for at least 72 h until a constant weight was achieved. The dry biomass of each component was recorded, and the proportions were calculated. All samples were ground to pass a 0.25 mm (60 mesh) sieve and stored for NSC analysis. The samples were analyzed for total soluble sugar and starch using the anthrone–sulfuric acid method (Yemm and Willis, 1954). About 0.1 g of the powder sample was placed in a 10-ml centrifuge tube with 5 ml 80% ethanol added. The mixture was incubated in a water bath at 80 °C for 30 min and then centrifuged at 4000 rpm for 10 min. The pellets were extracted twice more with 80% ethanol. The supernatants were pooled and stored at -20 °C for soluble sugar analysis. The pellets were saved for starch analysis. Total soluble sugar concentration was measured spectrophotometrically at 620 nm, using glucose as a standard. Starch was extracted from the saved pellets, as described by Li 21

Forest Ecology and Management 409 (2018) 19–28

D. Lu et al.

mean survival rate of Korean spruce was significantly higher than Manchurian walnut (Table 1).

et al. (2008). Starch concertation was also determined spectrophotometrically at 620 nm. The total NSC concentration was estimated as the sum of soluble sugar and starch in mg g−1 dry mass. The total NSC pool of each individual was calculated as the sum of the biomass of each organ multiplying its NSC concentration.

3.2. Seedling growth There was no significant interaction effect between canopy treatment and species on seedling height growth in 2015 (Table 1). Seedling height growth was not affected by canopy treatment and only differed between species (Table 1, Fig. 2A and B). At the end of the second growing season, canopy treatment and species showed a significant interaction effect (Table 1). The height of Manchurian walnut was significantly higher in larger gaps than in thinning plots and forest understories (P < .049, Fig. 2A). However, no height difference was found between treatments for Korean spruce (P > .932, Fig. 2B). Both species had similar RCD in 2015, and the RCD growth was not affected by canopy treatment (Table 1, Fig. 2C and D). In 2016, no significant interaction effect was found between canopy treatment and species on RCD growth. Both species had similar RCD, but RCD growth was significantly affected by canopy treatment (Table 1). Root collar diameter was higher in larger gaps than in forest understories (Fig. 2C and D). We did not find a significant interaction effect between canopy treatment and species on total seedling biomass in 2015 (Table 1). However, seedling biomass was significantly higher in larger gaps than in other treatments, and the mean biomass of Korean spruce was significantly higher than that of Manchurian walnut (Table 1, Fig. 3A and B). In 2016, there was a significant interaction effect between canopy treatment and species on total seedling biomass (Table 1). Seedling biomass of Manchurian walnut in larger gaps was more than two times of that in other treatments (P < .001, Fig. 3A). The biomass of Korean spruce was significantly higher in larger gaps than in thinning plots and forest understories (P < .003, Fig. 3B). The NSC concentration was only affected by seedling species during each year (Table 1). The mean NSC concentration of Manchurian walnut was significantly higher than that of Korean spruce (Table 1, Fig. 4A and B). Thus, the trends of NSC pool of two species in all treatments were similar with that of total seedling biomass. There was no significant interaction effect between canopy treatment and species on NSC pool in 2015, but NSC pool was affected by canopy treatment (Table 1). Both species had the highest NSC pool in larger gaps (Fig. 5A and B). In 2016, a significant interaction effect was found

2.4. Data analysis We averaged the data of all sub-plots in each larger gap and smaller gap treatment to represent the mean conditions of seedling survival and growth within the gaps before analysis. Two-way ANOVAs were used to test the effects of canopy treatment and species on seedling survival and growth (height, RCD, total biomass, NSC pool, and NSC concentration) at the end of each growing season (2015 and 2016). Two-way ANOVAs were also used to test the effects of canopy treatment and species on seedling biomass allocation and NSC pool allocation. Tukey’s post hoc tests were used to further examine the differences between treatment levels. Correlation analyses with a Pearson coefficient were conducted using all treatment data at the end of the second growing season to test the relationships between seedling survival and each growth characteristic (height, RCD, total biomass, NSC pool, and NSC concentration) for each species. Data were checked for normality and homogeneity before analysis. Log or square-root transformations were used when necessary to meet the statistical requirements. A P-value less than 0.05 was regarded as statistically significant. All the analyses were performed with R version 3.3.2 (R Core Team, 2016). Car package (Fox and Weisberg, 2011) was used for normality tests and ANOVAs. Lsmeans package (Lenth, 2016) was used for Tukey’s post hoc tests. 3. Results 3.1. Seedling survival Two-way ANOVA showed a significant interaction effect between canopy treatment and species on seedling survival in 2015 (Table 1). The survival rates of Manchurian walnut decreased from larger gaps to forest understories (Fig. 1A), but Korean spruce had similar survival rates in all treatments (P > .067, Fig. 1B). There was no significant interaction effect between canopy treatment and species in 2016 (Table 1). Both species had the highest survival rates in gaps and the lowest survival rates in forest understories (Fig. 1A and B), but the

Table 1 Effects of canopy treatment and species on seedling performance. Significant effects (P < .05) are bolded. Response variable

Effect

2015

2016

df

F-ratio

P-value

df

F-ratio

P-value

Survival rate

Treatment Species Treatment × Species

3 1 3

18.93 59.02 4.36

< .001 < .001 .015

3 1 3

16.00 20.98 0.94

< .001 < .001 .440

Height

Treatment Species Treatment × Species

3 1 3

0.09 161.65 0.49

.965 < .001 .693

3 1 3

10.60 148.77 5.18

< .001 < .001 .007

Root collar diameter

Treatment Species Treatment × Species

3 1 3

1.29 1.08 1.83

.303 .310 .172

3 1 3

12.01 2.10 3.02

< .001 .161 .051

Total biomass

Treatment Species Treatment × Species

3 1 3

18.80 245.45 2.08

< .001 < .001 .132

3 1 3

71.79 9.55 17.74

< .001 .005 < .001

NSC pool

Treatment Species Treatment × Species

3 1 3

12.32 0.43 1.21

< .001 .521 .330

3 1 3

83.81 124.54 54.46

< .001 < .001 < .001

NSC concentration

Treatment Species Treatment × Species

3 1 3

2.53 350.76 2.96

.084 < .001 .055

3 1 3

0.19 888.58 1.38

.903 < .001 .274

22

Forest Ecology and Management 409 (2018) 19–28

D. Lu et al.

Fig. 1. Seedling survival in different canopy treatments for two years. Different letters indicate significant differences (P < .05) among treatments. LG: larger gap; SG: smaller gap; TH: thinning; CK: control, forest understory.

Fig. 2. Seedling height and root collar diameter in different canopy treatments for two years. Different letters indicate significant differences (P < .05) among treatments. LG: larger gap; SG: smaller gap; TH: thinning; CK: control, forest understory.

23

Forest Ecology and Management 409 (2018) 19–28

D. Lu et al.

Fig. 3. Total seedling biomass in different canopy treatments for two years. Different letters indicate significant differences (P < .05) among treatments. LG: larger gap; SG: smaller gap; TH: thinning; CK: control, forest understory.

and root of Manchurian walnut was significantly higher in larger gaps than in forest understories in 2015 (P < .002), but the difference disappeared in 2016 (P > .331).

between canopy treatment and species on NSC pool (Table 1). Similarly, the NSC pool of Manchurian walnut in larger gaps was much higher than other treatments (Fig. 5A). Although the NSC pool of Korean spruce was higher in larger gaps, it was not statistically significant (P > .231, Fig. 5B) when compared to other treatments.

3.4. Relationship between survival and growth The two species showed different relationships between final survival rates and growth characteristics (Table 3). For Manchurian walnut, there were significant positive correlations between seedling survival and height, total biomass, and total NSC pool (Table 3). For Korean spruce, only RCD had a significantly positive correlation with final seedling survival rates (Table 3).

3.3. Biomass and NSC pool allocation Two-way ANOVAs indicated that the allocation patterns of biomass and NSC pool in two years were mainly affected by species (Table 2). Manchurian walnut stored more biomass and NSC pool in root (underground biomass), but Korean spruce stored more biomass and NSC pool in leaf and stem (aboveground biomass). For Korean spruce, the biomass ratios and NSC pool ratios of different organs were not affected by canopy treatments (P > .606). For Manchurian walnut, the ratios of leaf and stem were significantly higher in larger gaps and smaller gaps than in forest understories in 2015 (P < .005) and in larger gaps than in other treatments in 2016 (P < .004). In addition, the ratio of leaf

4. Discussion Converting monoculture larch plantations to mixed stands can improve ecological functions and ecosystem resilience, which meet the goal of modern forest management (Mason and Zhu, 2014). Compared

Fig. 4. Non-structural carbohydrate (NSC) concentration of each species in different canopy treatments for two years. Different letters indicate significant differences (P < .05) among treatments. LG: larger gap; SG: smaller gap; TH: thinning; CK: control, forest understory.

24

Forest Ecology and Management 409 (2018) 19–28

D. Lu et al.

Fig. 5. Total non-structural carbohydrate (NSC) pool of each species in different canopy treatments for two years. Different letters indicate significant differences (P < .05) among treatments. LG: larger gap; SG: smaller gap; TH: thinning; CK: control, forest understory.

with clear cutting, low-intensity canopy treatments (e.g., thinning or creation of small canopy gaps) could avoid extreme environmental conditions (Man et al., 2009) and maintain the current ecological services (Knapp et al., 2013). Our study found that the survival and growth of the two planted seedling species benefitted from the low-intensity canopy treatments, although species-specific responses were also observed because of the contrasting shade tolerance.

Table 3 The relationship between seedling survival and each growth variable for the two studied species in 2016. Significant relationships (P < .05) are bolded. Growth variable

Height Root collar diameter Total biomass Non-structural carbohydrate pool Non-structural carbohydrate concentration

4.1. Seedling survival and growth The survival of light-demanding Manchurian walnut exhibited greater differences among the canopy treatments compared with shadetolerant Korean spruce. During the two-year study period, Manchurian walnut had decreased survival rates from larger gaps to forest understories, and the differences were significant between treatments and control. These results were consistent with the widely published studies that reported that increased light intensity could promote the survival of light-demanding species (Zhang et al., 2013; Zhu et al., 2014). For example, the survival rates of Korean pine and Mongolian oak significantly increased from 20% light intensity to full light intensity (Zhang et al., 2013). The positive gap effects on seedling survival could last more than ten years for some light-demanding species, such as jack pine (Pinus banksiana Lamb.) (Man et al., 2009). For Korean spruce, seedlings also had higher survival rates in gaps than in forest

Juglans mandshurica

Picea koraiensis

r

P-value

r

P-value

0.70 0.51 0.58 0.57 0.28

0.003 0.053 0.025 0.026 0.315

0.39 0.62 0.47 0.43 −0.27

0.149 0.014 0.074 0.113 0.337

understories at the end of the study period. The results indicated that increased canopy openness could also promote the survival of shadetolerant species, although the trend was not as obvious as with lightdemanding species (Zhu et al., 2014). A previous study reported that coniferous shade-tolerant species could retain higher survival rates in gaps much larger than in our study (e.g., 1.0 ha) (Newsome et al., 2016). These results together indicated that shade-tolerant species might have the ability to survive in gaps with a wide size range. Our two studied species showed different growth patterns over time. At the end of the first growing season, both species had similar height and RCD between canopy treatments, but the total biomass in larger

Table 2 Effects of canopy treatment and seedling species on biomass allocation and NSC pool allocation (in parentheses). Significant effects (P < .05) are bolded. Response variable

Aboveground:Underground

Leaf:Stem

Leaf:Root

Stem:Root

Effect

Treatment Species Treatment × Species Treatment Species Treatment × Species Treatment Species Treatment × Species Treatment Species Treatment × Species

2015

2016

df

F-ratio

3 1 3 3 1 3 3 1 3 3 1 3

2.11 1380.78 1.32 6.43 298.54 4.22 6.07 1574.57 1.59 0.11 453.62 2.81

P-value (2.14) (408.71) (0.14) (1.89) (328.67) (1.77) (2.41) (698.69) (0.05) (1.43) (105.95) (0.55)

25

.128 < .001 .295 .003 < .001 .017 .004 < .001 .220 .954 < .001 .063

(.125) (< .001) (.932) (.161) (< .001) (.181) (.094) (< .001) (.984) (.261) (< .001) (.652)

df

F-ratio

3 1 3 3 1 3 3 1 3 3 1 3

0.69 548.04 0.72 6.55 130.17 4.35 1.20 275.76 0.54 1.76 599.40 2.88

P-value (0.34) (256.02) (0.31) (1.52) (140.64) (2.02) (0.12) (188.49) (0.21) (1.14) (218.87) (1.42)

.568 < .001 .550 .002 < .001 .015 .332 < .001 .661 .185 < .001 .059

(.800) (< .001) (.819) (.238) (< .001) (.140) (.945) (< .001) (.886) (.355) (< .001) (.263)

Forest Ecology and Management 409 (2018) 19–28

D. Lu et al.

larger gaps than in smaller gaps, creating any gaps with their sizes under the lower size limit of Zhu et al. (2015) or thinning with low intensities may not significantly promote the regeneration of light-demanding species. However, any gaps above the lower size limit could significantly promote the establishment of light-demanding species, which serves as direct evidence for biological or ecological support to the lower gap size limit proposed by Zhu et al. (2015). Compared with Manchurian walnut, increased canopy openness did not promote the height growth of Korean spruce. The height of Manchurian walnut was higher than that of Korean spruce before planting, and the difference became larger after two years even though Korean spruce was two years older, which indicated that height growth might be shade-tolerance related or species-specific. However, the RCD of Korean spruce was larger in larger gaps and smaller gaps than in thinning plots and forest understories. Our results were similar to a former study which reported that the root collar growth of white spruce and jack pine responded more sensitively to canopy treatment than height growth (Man et al., 2009). The total seedling biomass increased gradually from thinning plots to larger gaps, and the trend was generally consistent with the increased light intensity. Shade-tolerant species could survive under intact forest canopy for years, and their high light use efficiency allows them to grow in low light environments (Canham et al., 1990). The gradual increase of Korean spruce biomass indicated the flexibility of the species to different light environments, although no difference was found between thinning plots and forest understories during the study period. A previous study reported that seedling size of two shade-tolerant species increased with gap size, but Abies amabilis Douglas ex J. Forbes responded much slower than Tsuga heterophylla (Raf.) Sarg. (Gray and Spies, 1996). These findings indicated that responses of seedling growth to gap size may not only relate to shade tolerance (Denslow, 1980) but also be species-specific (Beaudet and Messier, 1998).

gaps was significantly higher than in other treatments. These results were probably due to the combination effects of transplanting, drought and canopy treatment. Seedlings may experience water stress after transplanting, which increases the probability of mortality or reduces its growth rate (Grossnickle, 2005; Thomas, 2009). Under drought conditions, seedlings would accelerate root growth to achieve more water absorption (Grossnickle, 2005). During our study period, a drought event occurred during our transplanting in 2015. Compared with the historical normal (i.e., mean annual precipitation of 810 mm), the annual precipitation in 2015 was more than 20% lower (630 mm). However, according to the ratios of biomass and NSC pool allocation, we did not find a significantly higher root ratio in any treatment. One possible reason may be that the treatment effect on soil water availability was masked by the drought encountered. The higher leaf biomass ratio observed for Manchurian walnut in larger gaps likely resulted from more new branches and leaves than in other treatments. The initial biomass of Manchurian walnut (5.9 g without leaves) was much smaller than that of Korean spruce (13.5 g) because of the younger age. However, the mean biomass of Manchurian walnut in larger gaps (30.5 g) was not only 1.6 times higher than that in other treatments, but also higher than that of older Korean spruce in the same treatment (25.7 g) after two years. The great biomass changes during the study period indicate that an age difference did not mask the species difference and the treatment effect. Compared with many enrichment planting studies (Knapp et al., 2013), our gap size was relatively small. Smaller gaps have been demonstrated to play little effect on natural regeneration of light-demanding species because of limited light environment (Denslow, 1980; Arevalo and Fernandez-Palacios, 2007). The great performance of Manchurian walnut observed in our study probably benefited from the monopodial crown structure of larch stands, which allowed more light to penetrate (Bartemucci et al., 2002). The other reason may be the routine removal of the competitive vegetation (Schulze, 2008; Ouédraogo et al., 2014). Thinning or gap creation could significantly increase the density and abundance of herbs and lead to a resource competition with target species (Kern et al., 2013; Haughian and Frego, 2016). Man et al. (2009) reported that a single weeding after partial cutting could remarkably promote the survival and growth of planted seedlings. Schulze (2008) also reported that the advantage of seedling performance would disappear after a few years if competitive species were not removed. We weeded monthly during each growing season (five times) because our previous enrichment planting experiment failed during the first year due to a lack of competition removal. The cost of weeding was about 25 USD each time for all sub-plots, which was acceptable in our study. Weeding was also widely applied after enrichment planting in Amazon forests and African forests due to the low costs (Schulze, 2008; Doucet et al., 2009). However, only a few competitors were found in such a weeding frequency. We suggest weeding less frequently (maybe twice a year) for a practical enrichment planting in larch plantations. The growth of Manchurian walnut in smaller gaps and thinning plots did not significantly differ from forest understories, even with the same practice of competition removal. Therefore, there may exist a lower limit of light intensity and duration for the growth of light-demanding species. For example, Cooper et al. (2014) inferred that direct radiation of several hours in the morning caused by gap creation could effectively promote oak (Quercus ithaburensis Decne.) sapling growth. In recent years, several studies quantified gap size limits in different climate regions in order to effectively differentiate forest gaps from natural canopy openings and open environments (Zhu et al., 2015; Schneider and Larson, 2017). According to the light intensity and average shadow length of gap border trees, Zhu et al. (2015) reported a lower gap size limit of 0.49 for expanded gaps and 0.23 for canopy gaps in terms of the ratio of gap diameter and border tree height. The ratios of the larger gaps and smaller gaps in our study are 0.71 and 0.38 for expanded gaps and 0.51 and 0.17 for canopy gaps, respectively. Given the significantly better growth performance of Manchurian walnut in

4.2. Biomass and NSC pool allocation Adaptive hypothesis claimed that species could allocate biomass according to the environment to achieve better resource use efficiently (Garnier, 1991). For example, some studies reported that plants would allocate more biomass to leaves or stems in shaded environments (Shipley and Meziane, 2002; Sevillano et al., 2016). However, Curt et al. (2005) claimed that different light environments did not affect the biomass allocation of European beech seedlings and the biomass allocation may be ontogenic. Our results only provided limited evidence for adaptive hypothesis, likely due to all of our treatments being limited to low-intensity canopy manipulation. Canopy treatments almost had no effects on biomass and NSC pool allocation during the two years, especially for Korean spruce. For Manchurian walnut, only leaf biomass ratio had a significant increase with increased canopy openness. Another possibility was that planted seedlings might need more time to show biomass allocation differences under different canopy treatments because of transplant shock. Thus, long-term monitoring may be needed to further clarify the treatment effect. Compared with canopy treatment effects, biomass and NSC pool allocation were consistently affected by species. Manchurian walnut allocated more biomass and NSC pool to root, while Korean spruce allocated more biomass and NSC pool to stem and leaf. These results were consistent with a former review, which reported that broadleaved deciduous seedlings stored more NSC in roots than coniferous seedlings (Villar-Salvador et al., 2015). However, we could rule out that the age difference might have resulted in different allocation patterns between the two species. Nevertheless, the different allocation strategies would optimize the resource utilization and contribute to the coexistence of these two species (Zhang et al., 2013).

26

Forest Ecology and Management 409 (2018) 19–28

D. Lu et al.

competitive shrubs and herbs could increase the survival and growth of planted seedlings and shorten the time to convert larch plantations to mixed stands. However, considering the costs of maintaining enrichment plantings, the appropriate times of weeding or other tending practices should be examined in the future. Our conversion recommendations are based on the performance of Manchurian walnut and Korean pine seedlings for the first two years. Continuously monitoring the experiment would be implemented to clarify the long-term effects of our treatments on larch plantation development. It is expected that additional silvicultural treatments would be required as the planted seedlings grow older or larger.

4.3. Relationship between survival and growth Positive relationships were found between survival and growth at the end of the second year. The survival of Korean spruce had the strongest positive relationship with RCD, while the survival of Manchurian walnut had the strongest positive relationship with seedling height. The different results were likely due to the contrasting shade tolerance of the two species (Beaudet and Messier, 1998). Beaudet and Messier (1998) reported that yellow birch (intermediate in shade tolerance) had higher height growth than sugar maple and beech (shade-tolerant) even in a shaded environment, and they regarded this phenomenon as a light-seeking strategy. However, some other studies reported that seedlings of different shade tolerance had high survival rates in all canopy treatments, but their growth increased with increasing canopy openness and showed species-specific differences (Löf et al., 2007; Sevillano et al., 2016). No matter positive relationship or no relationship, these findings suggested that the condition (e.g., light intensity) for seedling survival could be less rigorous than that for growth (Man et al., 2009). Seedling survival of both species had no relationship with NSC concentration. This finding was inconsistent with previous studies which reported that NSC concentration would decrease under shaded conditions (Myers and Kitajima, 2007). Compared with some extreme shade conditions (e.g., 0.08% full sunlight), larch plantations had better light environments. The mean PAR in forest understories was 28.8 μM m−2 s−1 during growing seasons, which was 30% light intensity of that in larger gaps. This elevated light condition in the understory prevented NSC concentration as a limiting factor to seedling survival.

Acknowledgements This work was financially supported by National Natural Science Foundation of China (31330016) and Key Research Program of Frontier Sciences, CAS (QYZDJ-SSW-DQC027). We thank Dr. Lizhong Yu and Dr. Yirong Sun from Institute of Applied Ecology, Chinese Academy of Sciences and Mr. Wenru Zhang from Dasuhe Forest Farm for their help during stand selection and seedling planting. We thank Miss Ting Zhang, Miss Jing Wang and Miss Ao Shen from Institute of Applied Ecology, Chinese Academy of Sciences for their help during sample measurements. We thank Dr. Min Zhang from Institute of Applied Ecology, Chinese Academy of Sciences for her suggestions on the statistical analysis. We thank the editor and two anonymous reviewers for their critical comments on the manuscript. We thank Miss Bridget Blood from Clemson University for language editing of the manuscript. References

5. Conclusions and management implications

Arevalo, J.R., Fernandez-Palacios, J.M., 2007. Treefall gaps and regeneration composition in the laurel forest of Anaga (Tenerife): a matter of size? Plant Ecol. 188, 133–143. Bartemucci, P., Coates, K.D., Harper, K.A., Wright, E.F., White, P., 2002. Gap disturbances in northern old-growth forests of British Columbia, Canada. J. Veg. Sci. 13, 685–696. Beaudet, M., Messier, C., 1998. Growth and morphological responses of yellow birch, sugar maple, and beech seedlings growing under a natural light gradient. Can. J. For. Res. 28, 1007–1015. Canham, C.D., Denslow, J.S., Platt, W.J., Runkle, J.R., Spies, T.A., White, P.S., 1990. Light regimes beneath closed canopies and tree-fall gaps in temperate and tropical forests. Can. J. For. Res. 20, 620–631. Chen, N.Q., Yin, J.D., 1990. Studies on effects of secondary regeneration of larch pure plantation under the successive rotations. In: The Proceedings of 2th Forest Silviculture Conference of Chinese Society of Forestry. Forestry Press, China, pp. 170–178. Chen, X., Li, B., 2003. Change in soil carbon and nutrient storage after human disturbance of a primary Korean pine forest in Northeast China. For. Ecol. Manage. 186, 197–206. Cooper, A., Moshe, Y., Zangi, E., Osem, Y., 2014. Are small-scale overstory gaps effective in promoting the development of regenerating oaks (Quercus ithaburensis) in the forest understory? New Forests 45, 843–857. Curt, T., Coll, L., Prévosto, B., Balandier, P., Kunstler, G., 2005. Plasticity in growth, biomass allocation and root morphology in beech seedlings as induced by irradiance and herbaceous competition. Ann. Forest Sci. 62, 51–60. Deal, R.L., 2007. Management strategies to increase stand structural diversity and enhance biodiversity in coastal rainforests of Alaska. Biol. Conserv. 137, 520–532. Denslow, J.S., 1980. Gap partitioning among tropical rainforest trees. Biotropica 12, 47–55. Doucet, J.L., Kouadio, Y.L., Monticelli, D., Lejeune, P., 2009. Enrichment of logging gaps with moabi (Baillonella toxisperma Pierre) in a Central African rain forest. For. Ecol. Manage. 258, 2407–2415. Fox, J., Weisberg S., 2011. An R Companion to Applied Regression, second ed. Sage, Thousand Oaks CA, < http://socserv.socsci.mcmaster.ca/jfox/Books/Companion > . Garnier, E., 1991. Resource capture, biomass allocation and growth in herbaceous plants. Trends Ecol. Evol. 6, 126–131. Gray, A.N., Spies, T.A., 1996. Gap size, within-gap position and canopy structure effects on conifer seedling establishment. J. Ecol. 84, 635–645. Grossnickle, S.C., 2005. Importance of root growth in overcoming planting stress. New Forests 30, 273–294. Haughian, S.R., Frego, K.A., 2016. Short-term effects of three commercial thinning treatments on diversity of understory vascular plants in white spruce plantations of northern New Brunswick. For. Ecol. Manage. 370, 45–55. Hu, H., Wang, G.G., Walke, J.L., Knapp, B.O., 2012. Silvicultural treatments for converting loblolly pine to longleaf pine dominance: Effects on planted longleaf pine seedlings. For. Ecol. Manage. 276, 209–216. Kern, C.C., D'Arnato, A.W., Strong, T.F., 2013. Diversifying the composition and structure of managed, late-successional forests with harvest gaps: What is the optimal gap size? For. Ecol. Manage. 304, 110–120.

The low-intensity canopy treatments examined in our study can be applied in a forest management plan that aims to gradually convert pure larch plantations to mixed stands. Because the survival and growth of the two studied species responded differently to canopy treatments, species-specific conversion approaches may be proposed. The growth of Manchurian walnut responded strongly to larger gaps while insensitively to other treatments. Therefore, if Manchurian walnut was selected for the conversion, larger gaps must be created. Because the larger gap tested in our study is only slightly above the lower gap size limit proposed by Zhu et al. (2015), we suggest creating gaps with a size above the lower size limit of Zhu et al. (2015) to establish Manchurian walnut or other light-demanding species in larch plantations. Unlike Manchurian walnut, Korean spruce gradually accumulated carbohydrate storage with the increase of canopy openness, indicating that the species could benefit from any incremental increase in canopy openness. Thus, if Korean spruce were selected, there would be no critical limit for gap size. Forest managers can plant Korean spruce with more flexibility. For example, Korean spruce could be planted in larch stands where tending practices are routinely conducted, although this would prolong the conversion period. Although we only planted two species in larch plantations, their performances also provided some reference for introducing other tree species. For example, different canopy treatments had little effect on NSC concentration, biomass allocation, and NSC pool allocation, which indicated that both tree species, regardless of shade tolerance, did not experience extreme environmental stress after the applied canopy treatments. Therefore, more planting species might be introduced to promote the conversion of larch plantations. Moreover, to facilitate the establishment of planted seedlings, competition removal or weeding is often applied during enrichment planting. Although the current study did not test the effects of weeding, our previous experiment and many published studies have demonstrated the importance of competition removal. Therefore, it is reasonable to assume that routine removal of 27

Forest Ecology and Management 409 (2018) 19–28

D. Lu et al.

Sevillano, I., Short, I., Grant, J., O’Reilly, C., 2016. Effects of light availability on morphology, growth and biomass allocation of Fagus sylvatica and Quercus robur seedlings. For. Ecol. Manage. 374, 11–19. Shipley, B., Meziane, D., 2002. The balanced-growth hypothesis and the allometry of leaf and root biomass allocation. Funct. Ecol. 16, 326–331. Thomas, D.S., 2009. Survival and growth of drought hardened Eucalyptus pilularis Sm. seedlings and vegetative cuttings. New Forests 38, 245–259. Thomas, S.C., Halpern, C.B., Falk, D.A., Liguori, D.A., Austin, K.A., 1999. Plant diversity in managed forests: understory responses to thinning and fertilization. Ecol. Appl. 9, 864–879. Van Hees, A., Clerkx, A., 2003. Shading and root–shoot relations in saplings of silver birch, pedunculate oak and beech. For. Ecol. Manage. 176, 439–448. Villar-Salvador, P., Uscola, M., Jacobs, D.F., 2015. The role of stored carbohydrates and nitrogen in the growth and stress tolerance of planted forest trees. New Forests 46, 813. Wang, C.Y., Liu, G.B., 2001. Spontaneous development of Larix olgensis plantations. For. Sci. Technol. 26, 12–14 (in Chinese). Wang, R., Xu, T., Yu, L., Zhu, J., Li, X., 2013. Effects of land use types on surface water quality across an anthropogenic disturbance gradient in the upper reach of the Hun River, Northeast China. Environ. Monit. Assess. 185, 4141–4151. Wang, J., Yan, Q., Yan, T., Song, Y., Sun, Y., Zhu, J., 2017. Rodent-mediated seed dispersal of Juglans mandshurica regulated by gap size and within-gap position in larch plantations: Implication for converting pure larch plantations into larch-walnut mixed forests. For. Ecol. Manage. 404, 205–213. Wang, Q., Wang, S., 2007. Soil organic matter under different forest types in Southern China. Geoderma 142, 349–356. Wang, Z., Zhang, S.Y., 1990. In: Larch forest in China. Forestry Publication House in China, Beijing, pp. 185–186 (in Chinese). Xu, M., Qi, Y., Gong, P., 2000. China's new forest policy. Science 289, 2049–2050. Yan, T., Zhu, J., Yang, K., Yu, L., Zhang, J., 2017. Nutrient removal under different harvesting scenarios for larch plantations in northeast China: implications for nutrient conservation and management. For. Ecol. Manage. 400, 150–158. Yan, Q., Zhu, J., Gang, Q., 2013. Comparison of spatial patterns of soil seed banks between larch plantations and adjacent secondary forests in Northeast China: implication for spatial distribution of larch plantations. Trees 27, 1747–1754. Yang, K., Shi, W., Zhu, J.J., 2013. The impact of secondary forests conversion into larch plantations on soil chemical and microbiological properties. Plant Soil 368, 535–546. Yang, K., Zhu, J.J., Yan, Q.L., Sun, J.X., 2010. Changes in soil P chemistry as affected by conversion of natural secondary forests to larch plantations. For. Ecol. Manage. 260, 422–428. Yemm, E., Willis, A., 1954. The estimation of carbohydrates in plant extracts by anthrone. Biochem. J. 57, 508–514. York, R.A., Heald, R.C., Battles, J.J., York, J.D., 2004. Group selection management in conifer forests: relationships between opening size and tree growth. Can. J. For. Res. 34, 630–641. Zhang, M., Zhu, J.J., Li, M.C., Zhang, G.Q., Yan, Q.L., 2013. Different light acclimation strategies of two coexisting tree species seedlings in a temperate secondary forest along five natural light levels. For. Ecol. Manage. 306, 234–242. Zhu, J., Yang, K., Yan, Q., Liu, Z., Yu, L., Wang, H., 2010. Feasibility of implementing thinning in even-aged Larix olgensis plantations to develop uneven-aged larch–broadleaved mixed forests. J. Forest Res. Jpn. 15, 71–80. Zhu, J., Zhang, G., Wang, G.G., Yan, Q., Lu, D., Li, X., Zheng, X., 2015. On the size of forest gaps: Can their lower and upper limits be objectively defined? Agric. Forest Meteorol. 213, 64–76. Zhu, J.J., Liu, Z.G., Wang, H.X., Yan, Q.L., Fang, H.Y., Hu, L.L., Yu, L.Z., 2008. Effects of site preparation on emergence and early establishment of Larix olgensis in montane regions of northeastern China. New Forests 36, 247–260. Zhu, J.J., Lu, D.L., Zhang, W.D., 2014. Effects of gaps on regeneration of woody plants: a meta-analysis. J. For. Res. 25, 501–510. Zhu, J.J., Mao, Z.H., Hu, L.L., Zhang, J.X., 2007. Plant diversity of secondary forests in response to anthropogenic disturbance levels in montane regions of northeastern China. J. For. Res. Jpn. 12, 403–416.

Knapp, B.O., Wang, G.G., Walker, J.L., 2013. Effects of canopy structure and cultural treatments on the survival and growth of Pinus palustris Mill. seedlings underplanted in Pinus taeda L. stands. Ecol. Eng. 57, 46–56. Kobe, R.K., 1997. Carbohydrate allocation to storage as a basis of interspecific variation in sapling survivorship and growth. Oikos 226–233. Kobe, R.K., Coates, K.D., 1997. Models of sapling mortality as a function of growth to characterize interspecific variation in shade tolerance of eight tree species of northwestern British Columbia. Can. J. For. Res. 27, 227–236. Lei, X.D., Zhang, H.R., Li, D.L., Liang, Y.Z., 2003. Comparison of plant species diversity of four forest types in over-logged forest area in Northeastern China. Chin. J. Ecol. 22, 47–50 (in Chinese with English abstract). Lenth, R.V., 2016. Least-squares means: the R package Lsmeans. J. Stat. Softw. 69, 1–33. Li, C., Zhou, X., 2000. Status and future trends in plantation silviculture in China. AMBIO: J. Hum. Environ. 29, 354–355. Li, M.H., Xiao, W.F., Wang, S.G., Cheng, G.W., Cherubini, P., Cai, X.H., Liu, X.L., Wang, X.D., Zhu, W.Z., 2008. Mobile carbohydrates in Himalayan treeline trees I. Evidence for carbon gain limitation but not for growth limitation. Tree Physiol. 28, 1287–1296. Li, W., 2004. Degradation and restoration of forest ecosystems in China. For. Ecol. Manage. 201, 33–41. Liu, Q., 1997. Structure and dynamics of the subalpine coniferous forest on Changbai Mountain, China. Plant Ecol. 132, 97–105. Liu, Z., Zhu, J., Hu, L., Wang, H., Mao, Z., Li, X., Zhang, L., 2005. Effects of thinning on microsites and natural regeneration in a Larix olgensis plantation in mountainous regions of eastern Liaoning Province, China. J. For. Res. 16, 193–199. Löf, M., Karlsson, M., Sonesson, K., Welander, T.N., Collet, C., 2007. Growth and mortality in underplanted tree seedlings in response to variations in canopy closure of Norway spruce stands. Forestry 80, 371–383. Man, R., Rice, J.A., MacDonald, G.B., 2009. Long-term response of planted conifers, natural regeneration, and vegetation to harvesting, scalping, and weeding on a boreal mixedwood site. For. Ecol. Manage. 258, 1225–1234. Mason, W.L., Zhu, J.J., 2014. Silviculture of planted forests managed for multi-functional objectives: lessons from Chinese and British experiences. In: Challenges and Opportunities for the World's Forests in the 21st Century. Springer, pp. 37–54. Myers, J.A., Kitajima, K., 2007. Carbohydrate storage enhances seedling shade and stress tolerance in a Neotropical forest. J. Ecol. 95, 383–395. Newsome, T.A., Brown, K.R., Nemec, A.F.L., 2016. Effects of opening size and microsite on performance of planted tree seedlings in high-elevation Engelmann spruce–subalpine fir forests managed as mountain caribou habitat in British Columbia. For. Ecol. Manage. 370, 31–44. Obrien, M.J., Leuzinger, S., Philipson, C.D., Tay, J., Hector, A., 2014. Drought survival of tropical tree seedlings enhanced by non-structural carbohydrate levels. Nat. Clim. Change 4, 710–714. Ouédraogo, D.Y., Fayolle, A., Daïnou, K., Demaret, C., Bourland, N., Lagoute, P., Doucet, J.L., 2014. Enrichment of logging gaps with a high conservation value species (Pericopsis elata) in a central African moist forest. Forests 5, 3031–3047. Owari, T., Tatsumi, S., Ning, L., Yin, M., 2015. Height growth of Korean pine seedlings planted under strip-cut larch plantations in Northeast China. Int. J. Forestry Res. 2015, 178681. Paquette, A., Bouchard, A., Cogliastro, A., 2006. Survival and growth of under-planted trees: a meta-analysis across four biomes. Ecol. Appl. 16, 1575–1589. R Core Team, 2016. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, < https://www.R-project. org/ > . Roberts, M.R., 2004. Response of the herbaceous layer to natural disturbance in North American forests. Can. J. Bot. 82, 1273–1283. Sala, A., Woodruff, D.R., Meinzer, F.C., 2012. Carbon dynamics in trees: feast or famine? Tree Physiol. 32, 764–775. Schneider, E.E., Larson, A.J., 2017. Spatial aspects of structural complexity in Sitka spruce–western hemlock forests, including evaluation of a new canopy gap delineation method. Can. J. For. Res. 1033–1044. Schulze, M., 2008. Technical and financial analysis of enrichment planting in logging gaps as a potential component of forest management in the eastern Amazon. For. Ecol. Manage. 255, 866–879.

28