Seedling survival within forest gaps: the effects of gap size, within-gap position and forest type on species of contrasting shade-tolerance in Northeast China

Seedling survival within forest gaps: the effects of gap size, within-gap position and forest... Abstract Previous studies on gap regeneration have seldom addressed potential differences due to within-gap position, condition of forest type and within-gap vegetation competition. We excluded understorey vegetation competition, and tested the effects of gap size, within-gap position and forest type on the survival of three species with contrasting shade tolerance by planting seedlings in twelve gaps (four each of small, medium and large size) created in secondary forests and four small gaps created in larch (Larix spp.) plantations. In the absence of vegetation competition, the survival of planted seedlings was generally higher in gaps than in forest understoreys despite there being significant differences among species. Light-demanding Juglans mandshurica survived better in large and medium gaps than in small gaps, and in the gap centre, transition, north edge and east edge than in other within-gap positions. Moreover, seedlings in larch plantation gaps had higher survival rates than in secondary forest gaps of the same size. Shade-tolerant Picea koraiensis had high survival rates that were not affected by gap size, within-gap position and forest type. Shade-tolerant Acer mono had extremely low survival rates in all treatments. Our findings suggest that, in addition to gap size, within-gap position and forest type condition should be considered to promote seedling establishment of light-demanding species during enrichment planting within gaps. Introduction Forest gaps, often resulting from a small-scale disturbance, are commonly found in forests throughout the world (Brokaw, 1987; Runkle and Yetter, 1987; McCarthy, 2001). Ecological effects of forest gaps have been extensively studied in tropical (Brokaw, 1987; Schnitzer and Carson, 2001), temperate (Runkle and Yetter, 1987; Weber et al., 2014) and boreal (de Romer et al., 2007) forests. Generally, gaps increased seedling or sapling density or diversity and helped maintain the stability of forest ecosystems (Adamic et al., 2016), although insignificant or even negative gap effects on tree regeneration were also reported (Hubbell et al., 1999; Arevalo and Fernandez-Palacios, 2007). However, the distributions of natural gaps are complex (Curzon and Keeton, 2010; Lobo and Dalling, 2013), and gap characteristics, such as size, age and shape, vary greatly (Vepakomma et al., 2012). In addition, the conditions of the stand where gaps are located vary considerably (Lobo and Dalling, 2013). All these uncertainties may limit the generality of findings from gap studies and constrain the application of findings in forest management. As the concept of close-to-nature silviculture develops, emulating natural disturbances becomes popular in modern forest management. One approach is to create artificial gaps based on stand conditions to achieve forest regeneration and other management objectives (Bieling, 2004; Streit et al., 2009). For example, slit-shaped gaps as a silvicultural technique were successful in increasing the seedling density of Picea abies (L.) Karst. in the Swiss Alps (Streit et al., 2009). In the northeast US, yellow birch (Betula alleghaniensis Britt.) was able to grow to the canopy in hemlock (Tsuga canadensis L.) gaps created by group selection (Webster and Lorimer, 2005). However, natural regeneration in gaps may depend on chance events (Brokaw and Busing, 2000), and desired tree species rarely appear without additional silvicultural interventions (Dauber et al., 2005). For example, a gap creation experiment did not increase the density of target tree species after 5 years (Madsen and Hahn, 2008), as many factors (e.g. lacking seed resource) may potentially weaken or invalidate the gap effects on natural regeneration. Enrichment planting (i.e. planting trees in forest stand gaps to increase the percentage of desired species or biodiversity) can increase the establishment of target species in gaps, which can be especially important for degraded forests (d’Oliveira, 2000; Ouedraogo et al., 2014; Schwartz et al., 2017). Seeds sowed or seedlings planted within gaps had higher survival rates and diameter growth increments after a few years (Schulze, 2008; Doucet et al., 2009; Ouedraogo et al., 2014). However, most of these studies focused on gap centre areas, with the exception of Gray and Spies (1996) and York et al. (2011), which may overestimate the gap effects on enrichment planting. The microenvironment within a gap is not homogeneous, especially in northern latitude forests (Canham et al., 1990; Brokaw and Busing, 2000; Diaci et al., 2008). Therefore, growth conditions at any given place within a gap would depend on the gap size and within-gap position. Furthermore, seedling survival and growth of desired species are often limited by competing vegetation within forest gaps, and the development of the competing vegetation within a gap also depends on the gap size and within-gap position. Evaluating gap centre effects or the whole gap effects would not reveal seedling performance among gaps of different sizes or different within-gap positions. For example, in the northern hemisphere, seedling survival may be higher in the southern part of gaps, where light intensity was modest but soil moisture was high (Gray and Spies, 1996; Wright et al., 1998). Although the gap partitioning hypothesis suggests that shade-tolerant species survive better in small gaps or forest understoreys compared with large gaps (Denslow, 1980), results from previous studies proved to be highly inconsistent (Wright et al., 1998; Zhu et al., 2014a; Sharma et al., 2016). We select secondary forests as our study object. Secondary forests are the largest ecosystem type in Northeast China, which function as important water conservation forests as well as timber reserves. Historically the main silvicultural approach focused on maximizing timber production, but it has been gradually changed to providing ecological service through close-to-nature silviculture such as gap creation. In this study, we eliminate vegetation competition and evaluate the effects of gap size, within-gap position, and forest type on the survival of three planted seedling species with different shade tolerance. Specifically, we address the following questions: (1) Can gaps increase the survival of planted seedlings compared with forest understoreys? (2) How do gap size, within-gap position and forest type affect seedling survival? (3) How does shade tolerance affect seedling survival? We hypothesize that gaps could promote seedling survival regardless of shade tolerance; light-demanding species may survive the highest in large gaps and gap centres, but shade-tolerant species would not show significant survival difference among gap sizes or within-gap positions. We expect that gap creation combined with enrichment planting and competition removal could be an appropriate approach for seedling establishment in the early stage. Methods Study site The experiment was carried out in Qingyuan Forest, the Chinese Ecosystem Research Network (CERN) site established by the Chinese Academy of Sciences (CAS). The study area is located within a mountain region in Liaoning Province, Northeast China (41°51′ N, 124°54′ E), with elevation ranging from 500 to 1100 m above sea level. It has a temperate monsoon climate, characterized by a warm and humid summer, and a cold and dry winter. The mean annual air temperature is 4.7°C. The coldest month is January, with a mean air temperature of −12.1°C, and the warmest month is July, with a mean air temperature of 21.0°C. The annual precipitation is 811 mm, and 80 per cent of precipitation falls during the summer. The frost-free period is 130 days, with an early frost occurring in October and a late frost occurring in April. The study area was characterized by mixed broadleaved-Korean pine (Pinus koraiensis Sieb. et Zucc.) forests, which were the largest natural forest community in Northeast China (Chen et al., 1994; Zhu et al., 2008). However, more than 70 per cent of these forests were severely degraded with almost no remaining natural Korean pine trees after a century’s timber exploitation (Chen et al., 1994; Zhu et al., 2008). Deciduous natural secondary forests and non-contiguous patches of larch (Larix spp.) plantations are currently dominating the study area. The dominant species in the secondary forests include ash (Fraxinus rhynchophylla Hance), Manchurian walnut (Juglans mandshurica Maxim.), Mongolian oak (Quercus mongolica Fisch.) and painted maple (Acer mono Maxim.) (Zhu et al., 2014b). The main plantation species are larch, Korean pine and Korean spruce (Picea koraiensis Nakai). Forests in this area are mainly disturbed by winds. Snowstorms and floods occur infrequently. Forest fires are rare. Disease, insect and browse are not common issues for the local species. Gap creation Overall, 16 gaps were artificially created in March 2015 (Table 1): 12 gaps in secondary forest stands with three size classes (large gaps: 658–984 m2, medium gaps: 476–512 m2 and small gaps: 176–260 m2), and four small gaps in larch plantation stands (130–206 m2). The mean ratio of gap diameter to gap border tree height (RD/H) of large, medium and small gaps in secondary forests is 1.66, 1.36 and 0.92, respectively. The mean RD/H of small gaps in larch plantations is 0.71. No large or medium gaps were created in larch plantations because a cutting area >200 m2 is prohibited before harvesting by the local Forestry Administration Agency. All gap locations were randomly selected in the forests with generally similar topography, slope and aspect. During the gap creation, all trees and shrubs taller than 30 cm within the gaps were cut off and removed. Every gap border tree was tagged, and its species, height and diameter at breast height were recorded. To minimize the negative effects of gap creation on forest soil and other trees, we cut down trees with chainsaws and moved them out of the gaps manually instead of using machines. The expanded gap size and shape were measured with a total station (TKS-202, China) as follows: (1) select an estimated gap centre point; (2) measure the direction and distance between the gap centre point and one gap border tree, and scale the data to coordinate paper; (3) repeat the previous procedure until all gap border trees are measured; and (4) delineate the gap shape, calculate the gap size and adjust the gap centre point. With the help of a total station, gaps could be created into any shape, and we approximated each gap consistently as a circle. Table 1 General description of the forest gaps created in March 2015 aForest gap Expanded gap size (m2) Altitude (m) Slope (°) Aspect (°) Mean height of gap border trees (m) Mean DBH of gap border trees (cm) bSpecies composition of gap border trees (%) L 984 678 22 191 20 27.3 QM (47), JM (24), FR (18) L 968 653 19 195 18 28.9 QM (77), FR (9), PD (9) L 688 679 25 170 20 30.8 QM (53), JM (47) L 658 672 25 220 20 32.6 QM (100) M 512 731 28 158 17 26.5 UP (35), JM (24), QM (18) M 492 762 26 170 19 32.5 QM (63), JM (19), FR (13) M 484 726 25 170 19 31.6 QM (47), JM (27), UP (20) M 476 664 18 168 19 28.8 QM (50), FM (25), BP (19) S 260 699 28 228 20 27.1 QM (77), LS (23) S 226 686 28 195 17 20.3 JM (63), FR (19), FM (13) S 184 708 22 208 17 21.1 FM (58), QM (25), FR (17) S 176 713 27 165 17 25.8 QM (75), FR (17), UP (8) S (Larch) 206 719 25 190 20 17 LS (100) S (Larch) 156 669 25 164 20 18.6 LS (100) S (Larch) 142 734 15 190 20 18.6 LS (100) S (Larch) 130 698 25 183 20 17.4 LS (100) aForest gap Expanded gap size (m2) Altitude (m) Slope (°) Aspect (°) Mean height of gap border trees (m) Mean DBH of gap border trees (cm) bSpecies composition of gap border trees (%) L 984 678 22 191 20 27.3 QM (47), JM (24), FR (18) L 968 653 19 195 18 28.9 QM (77), FR (9), PD (9) L 688 679 25 170 20 30.8 QM (53), JM (47) L 658 672 25 220 20 32.6 QM (100) M 512 731 28 158 17 26.5 UP (35), JM (24), QM (18) M 492 762 26 170 19 32.5 QM (63), JM (19), FR (13) M 484 726 25 170 19 31.6 QM (47), JM (27), UP (20) M 476 664 18 168 19 28.8 QM (50), FM (25), BP (19) S 260 699 28 228 20 27.1 QM (77), LS (23) S 226 686 28 195 17 20.3 JM (63), FR (19), FM (13) S 184 708 22 208 17 21.1 FM (58), QM (25), FR (17) S 176 713 27 165 17 25.8 QM (75), FR (17), UP (8) S (Larch) 206 719 25 190 20 17 LS (100) S (Larch) 156 669 25 164 20 18.6 LS (100) S (Larch) 142 734 15 190 20 18.6 LS (100) S (Larch) 130 698 25 183 20 17.4 LS (100) aL = large gap; M = medium gap; S = small gap; S (Larch) = small gap in larch plantations. bBP = Betula platyphylla Suk.; FM = Fraxinus mandshurica Rupr.; FR = Fraxinus rhynchophylla Hance; JM = Juglans mandshurica Maxim.; LS = Larix spp.; PD = Populus davidiana Dode; QM = Quercus mongolica Fisch. ex Ledeb.; UP = Ulmus pumila L. Table 1 General description of the forest gaps created in March 2015 aForest gap Expanded gap size (m2) Altitude (m) Slope (°) Aspect (°) Mean height of gap border trees (m) Mean DBH of gap border trees (cm) bSpecies composition of gap border trees (%) L 984 678 22 191 20 27.3 QM (47), JM (24), FR (18) L 968 653 19 195 18 28.9 QM (77), FR (9), PD (9) L 688 679 25 170 20 30.8 QM (53), JM (47) L 658 672 25 220 20 32.6 QM (100) M 512 731 28 158 17 26.5 UP (35), JM (24), QM (18) M 492 762 26 170 19 32.5 QM (63), JM (19), FR (13) M 484 726 25 170 19 31.6 QM (47), JM (27), UP (20) M 476 664 18 168 19 28.8 QM (50), FM (25), BP (19) S 260 699 28 228 20 27.1 QM (77), LS (23) S 226 686 28 195 17 20.3 JM (63), FR (19), FM (13) S 184 708 22 208 17 21.1 FM (58), QM (25), FR (17) S 176 713 27 165 17 25.8 QM (75), FR (17), UP (8) S (Larch) 206 719 25 190 20 17 LS (100) S (Larch) 156 669 25 164 20 18.6 LS (100) S (Larch) 142 734 15 190 20 18.6 LS (100) S (Larch) 130 698 25 183 20 17.4 LS (100) aForest gap Expanded gap size (m2) Altitude (m) Slope (°) Aspect (°) Mean height of gap border trees (m) Mean DBH of gap border trees (cm) bSpecies composition of gap border trees (%) L 984 678 22 191 20 27.3 QM (47), JM (24), FR (18) L 968 653 19 195 18 28.9 QM (77), FR (9), PD (9) L 688 679 25 170 20 30.8 QM (53), JM (47) L 658 672 25 220 20 32.6 QM (100) M 512 731 28 158 17 26.5 UP (35), JM (24), QM (18) M 492 762 26 170 19 32.5 QM (63), JM (19), FR (13) M 484 726 25 170 19 31.6 QM (47), JM (27), UP (20) M 476 664 18 168 19 28.8 QM (50), FM (25), BP (19) S 260 699 28 228 20 27.1 QM (77), LS (23) S 226 686 28 195 17 20.3 JM (63), FR (19), FM (13) S 184 708 22 208 17 21.1 FM (58), QM (25), FR (17) S 176 713 27 165 17 25.8 QM (75), FR (17), UP (8) S (Larch) 206 719 25 190 20 17 LS (100) S (Larch) 156 669 25 164 20 18.6 LS (100) S (Larch) 142 734 15 190 20 18.6 LS (100) S (Larch) 130 698 25 183 20 17.4 LS (100) aL = large gap; M = medium gap; S = small gap; S (Larch) = small gap in larch plantations. bBP = Betula platyphylla Suk.; FM = Fraxinus mandshurica Rupr.; FR = Fraxinus rhynchophylla Hance; JM = Juglans mandshurica Maxim.; LS = Larix spp.; PD = Populus davidiana Dode; QM = Quercus mongolica Fisch. ex Ledeb.; UP = Ulmus pumila L. Plot set up and seedling planting Each gap was divided into three areas with nine positions identified (Figure 1): gap centre (one position), gap transition (four positions: north, south, east and west), and gap edge (four positions: north, south, east and west). All positions were determined by their direction and relative distance from the gap centre. If the distance between gap centre point and each gap border tree was 1.0, the coordinates of 0.0, 0.5 and 1.0 on the north–centre–south line and east–centre–west line were the positions for gap centre, gap transition and gap edge, respectively. Figure 1 View largeDownload slide Schematic diagram of plot distribution within each gap. Figure 1 View largeDownload slide Schematic diagram of plot distribution within each gap. In each gap position, one 3 m × 3 m plot with a 0.2-m-wide buffer zone was set up for seedling planting. Eight plots were set up in the forest understoreys of secondary forests and larch plantations (four for each) as control. Totally, 152 plots were established in our experiment. All control plots were placed under the canopy without disturbance and more than 20 m distance from gaps, which assured the plot environment was not affected by gaps. Litter and woody debris within the plots resulted from the operation of gap creation were cleaned manually before planting. Competitive species (mainly herbs) in all plots including forest understoreys were removed monthly during two growing seasons after planting, because our previous experiment indicated that planted seedlings could be severely suppressed by herbs during the early stage, which could potentially negate the effect of gap size or within-gap position. Seedlings of three major species in Northeast China, painted maple, Manchurian walnut, and Korean spruce, were used for enrichment planting during late April and early May of 2015. Each plot was divided into six columns including 36 squares (0.5 m × 0.5 m), and each column was assigned one species randomly according to a random number sequence. For Manchurian walnut and Korean spruce, one seedling was planted in each square (12 seedlings per plot). For painted maple, two additional seedlings were planted in each square (36 seedlings per plot). As a result, a total of 5472 painted maple, 1824 Manchurian walnut and 1824 Korean spruce seedlings were planted. The number of painted maple seedlings was tripled because our early experiment found a relatively low survival rate of planted painted maple seedlings. Immediately after planting, seedlings of all plots were watered thoroughly because of abnormally dry soil conditions at the time of planting, although watering was not included in our original experimental design. All seedlings were bare-rooted seedlings, and each species was transplanted from the same local nursery. Painted maple and Manchurian walnut were 1-year-old seedlings, and Korean spruce were 3-year-old seedlings. During the 2-year study period, seedlings planted within each plot had maintained enough spacing to avoid competing with each other for light. Seedling measurement before planting In total, 12 seedlings of each species were selected and divided into three groups for measuring the seedling parameters before planting (Table 2). Height, basal diameter, biomass and non-structural carbohydrate (NSC) pool were measured. The NSC pool (in milligrams) was calculated as seedling biomass × NSC concentration, which was determined by anthrone colourimetry method (Li et al., 2008). Table 2 Seedling status before planting Species name Shade tolerance Age (y) aMean basal diameter (mm) bMean height (cm) cMean biomass (g) dMean NSC pool (mg) Acer mono Shade tolerant 1 3.31 (0.15) 21.80 (0.90) 1.69 (0.04) 46.98 (10.90) Juglans mandshurica Light demanding 1 6.32 (0.40) 31.00 (1.53) 5.85 (0.69) 439.24 (30.48) Picea koraiensis Shade tolerant 3 6.78 (0.61) 20.33 (0.88) 13.47 (0.33) 1121.21 (38.01) Species name Shade tolerance Age (y) aMean basal diameter (mm) bMean height (cm) cMean biomass (g) dMean NSC pool (mg) Acer mono Shade tolerant 1 3.31 (0.15) 21.80 (0.90) 1.69 (0.04) 46.98 (10.90) Juglans mandshurica Light demanding 1 6.32 (0.40) 31.00 (1.53) 5.85 (0.69) 439.24 (30.48) Picea koraiensis Shade tolerant 3 6.78 (0.61) 20.33 (0.88) 13.47 (0.33) 1121.21 (38.01) a,b,c,dValues in parenthesis represent standard error (N = 3). Table 2 Seedling status before planting Species name Shade tolerance Age (y) aMean basal diameter (mm) bMean height (cm) cMean biomass (g) dMean NSC pool (mg) Acer mono Shade tolerant 1 3.31 (0.15) 21.80 (0.90) 1.69 (0.04) 46.98 (10.90) Juglans mandshurica Light demanding 1 6.32 (0.40) 31.00 (1.53) 5.85 (0.69) 439.24 (30.48) Picea koraiensis Shade tolerant 3 6.78 (0.61) 20.33 (0.88) 13.47 (0.33) 1121.21 (38.01) Species name Shade tolerance Age (y) aMean basal diameter (mm) bMean height (cm) cMean biomass (g) dMean NSC pool (mg) Acer mono Shade tolerant 1 3.31 (0.15) 21.80 (0.90) 1.69 (0.04) 46.98 (10.90) Juglans mandshurica Light demanding 1 6.32 (0.40) 31.00 (1.53) 5.85 (0.69) 439.24 (30.48) Picea koraiensis Shade tolerant 3 6.78 (0.61) 20.33 (0.88) 13.47 (0.33) 1121.21 (38.01) a,b,c,dValues in parenthesis represent standard error (N = 3). Seedling survival and light environment monitoring Seedling survival was examined monthly for two growing seasons (2015 and 2016). If no green tissue on the seedling stem was observable (some seedlings were re-observed for 1 or 2 months), the seedling was regarded as dead (Obrien et al., 2014). The seedling survival rate of each species in each plot was calculated as follows: Seedlingsurvivalrate(%)=NtN0×100% (1) where, Nt is the number of seedlings that survived during the investigation at time, t, and N0 is the number of seedling initially planted. One gap of each size class and one control of each forest type were randomly selected for environmental monitoring. A data logger (WatchDog 1650 Micro Station, Spectrum Technologies, Inc., USA) was set up in each plot to continuously monitor photosynthetically active radiation (PAR) at 1 m above the ground, soil temperature at 5 cm below the ground, and soil moisture at 5 cm below the ground (Table 3). Table 3 Mean light, soil temperature and soil moisture environment in gaps and forest understoreys during two growing seasons Position Photosynthetically active radiation (PAR, μM m−2 s−1) Soil temperature at 5 cm (°C) Soil moisture at 5 cm (%) aLarge bMedium cSmall dSmall (Larch) Large Medium Small Small (Larch) Large Medium Small Small (Larch) Centre 174.53 141.39 90.31 123.83 18.67 18.44 18.38 17.82 29.75 27.09 17.43 20.79 North transition 179.71 125.79 94.31 155.87 19.45 18.38 18.82 18.58 19.13 12.44 16.60 25.92 East transition 150.43 108.50 75.90 93.02 19.57 18.98 17.85 17.61 15.37 15.19 18.81 30.47 West transition 148.45 127.68 62.76 121.61 18.40 17.68 18.07 18.11 32.17 21.98 20.97 19.36 South transition 119.74 91.61 41.65 69.84 17.81 16.99 17.41 17.04 28.18 41.48 11.80 24.07 North edge 67.18 43.59 28.51 102.96 18.33 17.39 16.46 18.89 13.76 15.40 13.50 9.18 East edge 84.86 28.06 41.82 61.08 18.63 17.51 17.51 17.55 9.04 17.19 19.63 12.00 West edge 54.75 50.03 31.01 94.97 17.63 17.19 17.10 17.91 13.91 14.23 10.70 9.90 South edge 54.73 35.96 21.10 53.43 17.45 16.78 16.83 17.10 25.43 35.99 15.68 11.23 eForest understorey 13.10 28.80 16.96 16.98 14.42 13.86 Position Photosynthetically active radiation (PAR, μM m−2 s−1) Soil temperature at 5 cm (°C) Soil moisture at 5 cm (%) aLarge bMedium cSmall dSmall (Larch) Large Medium Small Small (Larch) Large Medium Small Small (Larch) Centre 174.53 141.39 90.31 123.83 18.67 18.44 18.38 17.82 29.75 27.09 17.43 20.79 North transition 179.71 125.79 94.31 155.87 19.45 18.38 18.82 18.58 19.13 12.44 16.60 25.92 East transition 150.43 108.50 75.90 93.02 19.57 18.98 17.85 17.61 15.37 15.19 18.81 30.47 West transition 148.45 127.68 62.76 121.61 18.40 17.68 18.07 18.11 32.17 21.98 20.97 19.36 South transition 119.74 91.61 41.65 69.84 17.81 16.99 17.41 17.04 28.18 41.48 11.80 24.07 North edge 67.18 43.59 28.51 102.96 18.33 17.39 16.46 18.89 13.76 15.40 13.50 9.18 East edge 84.86 28.06 41.82 61.08 18.63 17.51 17.51 17.55 9.04 17.19 19.63 12.00 West edge 54.75 50.03 31.01 94.97 17.63 17.19 17.10 17.91 13.91 14.23 10.70 9.90 South edge 54.73 35.96 21.10 53.43 17.45 16.78 16.83 17.10 25.43 35.99 15.68 11.23 eForest understorey 13.10 28.80 16.96 16.98 14.42 13.86 a,b,cLarge, medium and small represent large gaps, medium gaps and small gaps in secondary forests, respectively. dSmall (Larch) represents small gaps in larch plantations. eForest understorey represents the control plots of each forest type. Table 3 Mean light, soil temperature and soil moisture environment in gaps and forest understoreys during two growing seasons Position Photosynthetically active radiation (PAR, μM m−2 s−1) Soil temperature at 5 cm (°C) Soil moisture at 5 cm (%) aLarge bMedium cSmall dSmall (Larch) Large Medium Small Small (Larch) Large Medium Small Small (Larch) Centre 174.53 141.39 90.31 123.83 18.67 18.44 18.38 17.82 29.75 27.09 17.43 20.79 North transition 179.71 125.79 94.31 155.87 19.45 18.38 18.82 18.58 19.13 12.44 16.60 25.92 East transition 150.43 108.50 75.90 93.02 19.57 18.98 17.85 17.61 15.37 15.19 18.81 30.47 West transition 148.45 127.68 62.76 121.61 18.40 17.68 18.07 18.11 32.17 21.98 20.97 19.36 South transition 119.74 91.61 41.65 69.84 17.81 16.99 17.41 17.04 28.18 41.48 11.80 24.07 North edge 67.18 43.59 28.51 102.96 18.33 17.39 16.46 18.89 13.76 15.40 13.50 9.18 East edge 84.86 28.06 41.82 61.08 18.63 17.51 17.51 17.55 9.04 17.19 19.63 12.00 West edge 54.75 50.03 31.01 94.97 17.63 17.19 17.10 17.91 13.91 14.23 10.70 9.90 South edge 54.73 35.96 21.10 53.43 17.45 16.78 16.83 17.10 25.43 35.99 15.68 11.23 eForest understorey 13.10 28.80 16.96 16.98 14.42 13.86 Position Photosynthetically active radiation (PAR, μM m−2 s−1) Soil temperature at 5 cm (°C) Soil moisture at 5 cm (%) aLarge bMedium cSmall dSmall (Larch) Large Medium Small Small (Larch) Large Medium Small Small (Larch) Centre 174.53 141.39 90.31 123.83 18.67 18.44 18.38 17.82 29.75 27.09 17.43 20.79 North transition 179.71 125.79 94.31 155.87 19.45 18.38 18.82 18.58 19.13 12.44 16.60 25.92 East transition 150.43 108.50 75.90 93.02 19.57 18.98 17.85 17.61 15.37 15.19 18.81 30.47 West transition 148.45 127.68 62.76 121.61 18.40 17.68 18.07 18.11 32.17 21.98 20.97 19.36 South transition 119.74 91.61 41.65 69.84 17.81 16.99 17.41 17.04 28.18 41.48 11.80 24.07 North edge 67.18 43.59 28.51 102.96 18.33 17.39 16.46 18.89 13.76 15.40 13.50 9.18 East edge 84.86 28.06 41.82 61.08 18.63 17.51 17.51 17.55 9.04 17.19 19.63 12.00 West edge 54.75 50.03 31.01 94.97 17.63 17.19 17.10 17.91 13.91 14.23 10.70 9.90 South edge 54.73 35.96 21.10 53.43 17.45 16.78 16.83 17.10 25.43 35.99 15.68 11.23 eForest understorey 13.10 28.80 16.96 16.98 14.42 13.86 a,b,cLarge, medium and small represent large gaps, medium gaps and small gaps in secondary forests, respectively. dSmall (Larch) represents small gaps in larch plantations. eForest understorey represents the control plots of each forest type. Data analysis Generalized linear mixed-effect models (logistic regressions with binomial distribution and logit link function) were used to compare the final seedling survival of each species between gaps and forest understoreys in secondary forests and larch plantations. Survival analysis models (Mixed-effect Cox regressions based on condition independence through time) were used for small gaps in two forest types to examine the effects of forest type and species on seedling survival. Similarly, survival analysis models were used for all secondary forest gaps to examine the effects of gap size and species on seedling survival. Finally, Logistic regressions were used to compare the final seedling survival of each species between different within-gap positions. For all the analyses, gap was regarded as a random effect. A P < 0.05 was regarded as statistically significant. All the analyses were performed with R version 3.3.2 (R Core Team, 2016). The coxme package (Therneau, 2018) was used for survival analyses. The lme4 package (Bates et al., 2015) was used for logistic regressions. Results Seedling survival in gaps and forest understoreys Logistic regressions indicated that the final seedling survival rates were significantly higher in gaps than forest understoreys in both secondary forests and larch plantations for Korean spruce (P < 0.05, Table 4) and Manchurian walnut (P < 0.001, Table 4). The final seedling survival rates of painted maple were significantly higher in secondary forest gaps than understoreys (P < 0.01, Table 4), but there was only a trend of significant survival difference between larch plantation gaps and understoreys (P = 0.058, Table 4). The mean final survival rates in gaps/understoreys were 61.4 per cent/45.0 per cent, 48.2 per cent/17.5 per cent and 8.0 per cent/1.4 per cent in secondary forests, and 62.8 per cent/42.5 per cent, 52.2 per cent/20.0 per cent and 9.8 per cent/4.9 per cent in larch plantations of Korean spruce, Manchurian walnut and painted maple, respectively. Table 4 Seedling survival estimates of logistic regressions for three planted species in larch plantations (4 gaps and 4 controls) and secondary forests (12 gaps and 4 controls). Significant effects (P < 0.05) are in bold Species Model term Larch plantation Secondary forest Coefficient SE Wald Z P-value Coefficient SE Wald Z P-value Picea koraiensis Understorey (intercept) −0.33 0.33 −0.98 0.329 −0.20 0.33 −0.62 0.537 Gap 0.85 0.36 2.38 0.018 0.67 0.33 2.03 0.042 Juglans mandshurica Understorey (intercept) −1.39 0.39 −3.51 <0.001 −1.81 0.45 −3.99 <0.001 Gap 1.48 0.41 3.61 <0.001 1.74 0.44 3.97 <0.001 Acer mono Understorey (intercept) −2.97 0.39 −7.67 <0.001 −4.36 0.72 −6.08 <0.001 Gap 0.75 0.40 1.89 0.058 1.89 0.72 2.64 0.008 Species Model term Larch plantation Secondary forest Coefficient SE Wald Z P-value Coefficient SE Wald Z P-value Picea koraiensis Understorey (intercept) −0.33 0.33 −0.98 0.329 −0.20 0.33 −0.62 0.537 Gap 0.85 0.36 2.38 0.018 0.67 0.33 2.03 0.042 Juglans mandshurica Understorey (intercept) −1.39 0.39 −3.51 <0.001 −1.81 0.45 −3.99 <0.001 Gap 1.48 0.41 3.61 <0.001 1.74 0.44 3.97 <0.001 Acer mono Understorey (intercept) −2.97 0.39 −7.67 <0.001 −4.36 0.72 −6.08 <0.001 Gap 0.75 0.40 1.89 0.058 1.89 0.72 2.64 0.008 Table 4 Seedling survival estimates of logistic regressions for three planted species in larch plantations (4 gaps and 4 controls) and secondary forests (12 gaps and 4 controls). Significant effects (P < 0.05) are in bold Species Model term Larch plantation Secondary forest Coefficient SE Wald Z P-value Coefficient SE Wald Z P-value Picea koraiensis Understorey (intercept) −0.33 0.33 −0.98 0.329 −0.20 0.33 −0.62 0.537 Gap 0.85 0.36 2.38 0.018 0.67 0.33 2.03 0.042 Juglans mandshurica Understorey (intercept) −1.39 0.39 −3.51 <0.001 −1.81 0.45 −3.99 <0.001 Gap 1.48 0.41 3.61 <0.001 1.74 0.44 3.97 <0.001 Acer mono Understorey (intercept) −2.97 0.39 −7.67 <0.001 −4.36 0.72 −6.08 <0.001 Gap 0.75 0.40 1.89 0.058 1.89 0.72 2.64 0.008 Species Model term Larch plantation Secondary forest Coefficient SE Wald Z P-value Coefficient SE Wald Z P-value Picea koraiensis Understorey (intercept) −0.33 0.33 −0.98 0.329 −0.20 0.33 −0.62 0.537 Gap 0.85 0.36 2.38 0.018 0.67 0.33 2.03 0.042 Juglans mandshurica Understorey (intercept) −1.39 0.39 −3.51 <0.001 −1.81 0.45 −3.99 <0.001 Gap 1.48 0.41 3.61 <0.001 1.74 0.44 3.97 <0.001 Acer mono Understorey (intercept) −2.97 0.39 −7.67 <0.001 −4.36 0.72 −6.08 <0.001 Gap 0.75 0.40 1.89 0.058 1.89 0.72 2.64 0.008 Seedling survival in secondary forest gaps of different sizes Cox regressions indicated that seedling survival was significantly affected by gap size and species during the study period (P < 0.001). Compared with small gaps, the mean hazard of mortality reduced 15 and 11 per cent in large and medium gaps, respectively (Table 5). Compared with painted maple, the mean hazard of mortality reduced 65 and 80 per cent for Manchurian walnut and Korean spruce, respectively (Table 5). Manchurian walnut had significantly higher survival rate in large and medium gaps than in small gaps (P < 0.001; Figure 2). No significant survival difference was found for Korean spruce between large, medium and small gaps (P > 0.60; Figure 2). Painted maple had significantly higher survival rate in large and medium gaps than in small gaps (P < 0.05; Figure 2). The rank of survival rates among species did not change during the study period (Figure 2). Korean spruce had the highest survival rate, and painted maple had the lowest survival rate with remarkable mortality occurring 30 days after planting (Figure 2). Table 5 Survival results of Cox regressions for three planted species in gaps of different sizes and forest types. The analysis of small gaps in two forest types includes 4320 seedlings with 943 seedlings surviving beyond the last census. The analysis of gaps of different sizes in secondary forests includes 6480 seedlings with 1493 seedlings surviving beyond the last census. Seedling survival was assessed monthly during each growing season (2015 and 2016). Significant effects (P < 0.05) are in bold Model term Coefficient dEXP (coefficient) SE (coefficient) Wald Z P-value Gaps of different sizes in secondary forests  aSize – medium −0.12 0.89 0.06 −2.14 0.032  Size – large −0.16 0.85 0.06 −2.99 0.003  bSpecies – JM −1.05 0.35 0.06 −16.59 <0.001  Species – PK −1.62 0.20 0.07 −22.16 <0.001  Size – medium: species – JM −0.32 0.73 0.10 −3.36 <0.001  Size – large: species – JM −0.39 0.68 0.10 −4.01 <0.001  Size – medium: species – PK −0.05 0.95 0.11 −0.47 0.630  Size – large: species – PK 0.04 1.05 0.11 0.43 0.670 Small gaps in two forest types  cForest type – Larch −0.14 0.87 0.05 −2.72 0.007  Species – JM −1.06 0.35 0.06 −16.63 <0.001  Species – PK −1.62 0.20 0.07 −22.15 <0.001  Forest type – Larch: species – JM −0.26 0.77 0.09 −2.80 0.005  Forest type – Larch: species – PK 0.01 1.01 0.10 0.08 0.940 Model term Coefficient dEXP (coefficient) SE (coefficient) Wald Z P-value Gaps of different sizes in secondary forests  aSize – medium −0.12 0.89 0.06 −2.14 0.032  Size – large −0.16 0.85 0.06 −2.99 0.003  bSpecies – JM −1.05 0.35 0.06 −16.59 <0.001  Species – PK −1.62 0.20 0.07 −22.16 <0.001  Size – medium: species – JM −0.32 0.73 0.10 −3.36 <0.001  Size – large: species – JM −0.39 0.68 0.10 −4.01 <0.001  Size – medium: species – PK −0.05 0.95 0.11 −0.47 0.630  Size – large: species – PK 0.04 1.05 0.11 0.43 0.670 Small gaps in two forest types  cForest type – Larch −0.14 0.87 0.05 −2.72 0.007  Species – JM −1.06 0.35 0.06 −16.63 <0.001  Species – PK −1.62 0.20 0.07 −22.15 <0.001  Forest type – Larch: species – JM −0.26 0.77 0.09 −2.80 0.005  Forest type – Larch: species – PK 0.01 1.01 0.10 0.08 0.940 a,b,cGap size – small, species – AM and forest type – secondary forest are baselines. dEXP (coefficient) indicates the mean hazard of mortality relative to baselines. Table 5 Survival results of Cox regressions for three planted species in gaps of different sizes and forest types. The analysis of small gaps in two forest types includes 4320 seedlings with 943 seedlings surviving beyond the last census. The analysis of gaps of different sizes in secondary forests includes 6480 seedlings with 1493 seedlings surviving beyond the last census. Seedling survival was assessed monthly during each growing season (2015 and 2016). Significant effects (P < 0.05) are in bold Model term Coefficient dEXP (coefficient) SE (coefficient) Wald Z P-value Gaps of different sizes in secondary forests  aSize – medium −0.12 0.89 0.06 −2.14 0.032  Size – large −0.16 0.85 0.06 −2.99 0.003  bSpecies – JM −1.05 0.35 0.06 −16.59 <0.001  Species – PK −1.62 0.20 0.07 −22.16 <0.001  Size – medium: species – JM −0.32 0.73 0.10 −3.36 <0.001  Size – large: species – JM −0.39 0.68 0.10 −4.01 <0.001  Size – medium: species – PK −0.05 0.95 0.11 −0.47 0.630  Size – large: species – PK 0.04 1.05 0.11 0.43 0.670 Small gaps in two forest types  cForest type – Larch −0.14 0.87 0.05 −2.72 0.007  Species – JM −1.06 0.35 0.06 −16.63 <0.001  Species – PK −1.62 0.20 0.07 −22.15 <0.001  Forest type – Larch: species – JM −0.26 0.77 0.09 −2.80 0.005  Forest type – Larch: species – PK 0.01 1.01 0.10 0.08 0.940 Model term Coefficient dEXP (coefficient) SE (coefficient) Wald Z P-value Gaps of different sizes in secondary forests  aSize – medium −0.12 0.89 0.06 −2.14 0.032  Size – large −0.16 0.85 0.06 −2.99 0.003  bSpecies – JM −1.05 0.35 0.06 −16.59 <0.001  Species – PK −1.62 0.20 0.07 −22.16 <0.001  Size – medium: species – JM −0.32 0.73 0.10 −3.36 <0.001  Size – large: species – JM −0.39 0.68 0.10 −4.01 <0.001  Size – medium: species – PK −0.05 0.95 0.11 −0.47 0.630  Size – large: species – PK 0.04 1.05 0.11 0.43 0.670 Small gaps in two forest types  cForest type – Larch −0.14 0.87 0.05 −2.72 0.007  Species – JM −1.06 0.35 0.06 −16.63 <0.001  Species – PK −1.62 0.20 0.07 −22.15 <0.001  Forest type – Larch: species – JM −0.26 0.77 0.09 −2.80 0.005  Forest type – Larch: species – PK 0.01 1.01 0.10 0.08 0.940 a,b,cGap size – small, species – AM and forest type – secondary forest are baselines. dEXP (coefficient) indicates the mean hazard of mortality relative to baselines. Figure 2 View largeDownload slide Survival rates in secondary forest gaps of different sizes (N = 4) for Acer mono (AM), Juglans mandshurica (JM) and Picea koraiensis (PK). Figure 2 View largeDownload slide Survival rates in secondary forest gaps of different sizes (N = 4) for Acer mono (AM), Juglans mandshurica (JM) and Picea koraiensis (PK). Logistic regressions indicated that Manchurian walnut finally had lower survival rates in west edge and south edge, but the survival rates in other within-gap positions were not significantly different from the gap centre (Figure 3). For Manchurian walnut, the survival patterns were similar in large, medium and small gaps (Figure 3). For Korean spruce, no significant survival differences were found among within-gap positions in large, medium and small gaps (Figure 3). The survival of painted maple within gaps showed random patterns. For example, seedling had lower survival rates in the west transition of small gaps, in the south edge and east edge of medium gaps, but no significant survival differences were found within large gaps (Figure 3). Figure 3 View largeDownload slide Final survival rates of nine within-gap positions (N = 4) for all gaps in larch plantations and secondary forests. Significant effects (P < 0.05; compared with the survival rate in gap centre which is regarded as the baseline) are in bold. Figure 3 View largeDownload slide Final survival rates of nine within-gap positions (N = 4) for all gaps in larch plantations and secondary forests. Significant effects (P < 0.05; compared with the survival rate in gap centre which is regarded as the baseline) are in bold. Seedling survival in small gaps of two forest types Cox regressions indicated that seedling survival was significantly affected by forest type and species during the study period (P < 0.001). Compared with secondary forests, the mean hazard of mortality reduced 13 per cent in larch plantations (Table 5). Compared with painted maple, the mean hazard of mortality reduced 65 and 80 per cent for Manchurian walnut and Korean spruce, respectively (Table 5). Manchurian walnut had significantly higher survival rates in larch plantation gaps than in secondary forest gaps (P < 0.01; Figure 4). No significant survival difference was found for Korean spruce between gaps in secondary forests and larch plantations (P > 0.90; Figure 4). Painted maple had significantly higher survival rates in larch plantation gaps than in secondary forest gaps (P < 0.01; Figure 4). The rank of seedling survival among species was similar in gaps of two forest types (Figure 4). Figure 4 View largeDownload slide Survival rates of small gaps in larch plantations (Lar; N = 4) and secondary forests (Sec; N = 4) for Acer mono (AM), Juglans mandshurica (JM) and Picea koraiensis (PK). Figure 4 View largeDownload slide Survival rates of small gaps in larch plantations (Lar; N = 4) and secondary forests (Sec; N = 4) for Acer mono (AM), Juglans mandshurica (JM) and Picea koraiensis (PK). Logistic regressions indicated that the final survival rates of Manchurian walnut were significantly lower in the west edge, south edge and east edge of larch plantation gaps, but the survival rates in other within-gap positions were not significantly different from the gap centre (Figure 3). The final survival patterns of Manchurian walnut were similar in two forest types (Figure 3). For Korean spruce, no significant survival differences were found among within-gap positions in larch plantation gaps, which was similar to the patterns in small secondary forest gaps (Figure 3). There was no significant survival difference among within-gap positions for painted maple in larch plantation gaps, which differed from small secondary forest gaps (Figure 3). Discussion Seedling survival in gaps and forest understoreys Enrichment planting in artificial gaps is an efficient way to promote seedling establishment in an existing stand. Gaps improve the microenvironment which affects nutrient release and seedling establishment (Muscolo et al., 2007; Liu et al., 2011; Vilhar et al., 2015). In our study, almost all species survived better in gaps than in forest understoreys regardless of shade tolerance. These findings supported that the regeneration of shade-tolerant species could also benefit from gaps (Wang and Liu, 2011). Although all studied species had positive responses to gaps compared with forest understoreys, the increase of seedling survival rates of the two shade-tolerant species (painted maple and Korean spruce) were lower than the light-demanding species (Manchurian walnut). These results were consistent with a recent meta-analysis, which reported that the relative density of light-demanding species in gaps was significantly higher than shade-tolerant species (Zhu et al., 2014a). Seedling survival in relation to gap sizes and within-gap positions Gap size is one of the most important factors affecting seedling regeneration within gaps (Bolton and D’Amato, 2011; Kern et al., 2012, 2016). However, previous studies indicated that the effect of gap size on regeneration was confounded by several factors (Bolton and D’Amato, 2011; Kern et al., 2016). For example, creating large gaps may reduce tree density due to the increased resource competition from other vegetation, such as shrubs and herbs (Kern et al., 2013). Therefore, we excluded the competition effect by manually removing non-target species from planting plots to further examine seedling responses to gaps. In addition, the environment within gaps is spatially heterogeneous, especially at high latitudes (Canham et al., 1990; Heinemann et al., 2000). Our study was carried out in the Northern Hemisphere, and all gaps had general south aspects, which may further amplify the environmental differences among within-gap positions, especially along the north–centre–south line within gaps. Therefore, it was necessary to examine the effects of within-gap positions on seedling survival. For light-demanding Manchurian walnut, the survival rate decreased faster in small gaps compared with large and medium gaps over the 2-year study period. Seedlings planted in the south edge and west edge usually had lower survival rates than those in other within-gap positions at the end of the second growing season, which was generally consistent with the light distribution within gaps (Canham et al., 1990; Diaci et al., 2008). These results indicated that the survival of light-demanding species benefited from the higher light supply in larger gaps or within-gap positions with more exposure (Galhidy et al., 2006; Cooper et al., 2014a, 2014b). However, Knapp et al. (2013) found that the mortality of planted Pinus palustris Mill. seedlings was higher in the gap centre than gap edge, and in the northern part compared with the southern part of gaps, but they attributed their results to climatic drought during the study period. A global meta-analysis indicated that annual seedling survival rates could differ significantly throughout different biomes even under the same canopy treatment (Paquette et al., 2006). Therefore, the effects of gap size and within-gap position on seedling establishment must be interpreted with the context and limitation of each particular study. Although shade-tolerant Korean spruce had higher survival rates in gaps than in forest understoreys, similar survival rates were found in gaps of different sizes and in different within-gap positions over the 2 years. A previous study focusing on regeneration of shade-tolerant coniferous species within gaps also found higher seedling survival rates in gaps than in forest understoreys, but all the species were more abundant in the shaded parts within gaps (Gray and Spies, 1996). These findings indicated that shade-tolerant coniferous species, such as Korean spruce, could benefit from increased light intensity only within a limited range, but may not partition along gap resource gradient (e.g. light intensity) as shown in light-demanding species. Previous studies reported that shade-tolerant painted maple could survive in forest understoreys due to high photosynthetic rate at low light environment (Kitao et al., 2006). However, the mean final survival rates of painted maple were generally <10 per cent in our study. Although significant survival differences existed between forest types, gap sizes and within-gap positions, the extremely low survival rates of painted maple had little meaning in practice. A previous study monitored the survival of three maple species (Acer pensylvanicum, Acer rubrum and Acer saccharum) planted in gaps and reported a mean survival rate of 79 per cent after three years (Sipe and Bazzaz, 1995), which is much higher than our results. The low survival rates of painted maple may be caused by the much smaller seedling dimensions, biomass and NSC pool at the time of planting. We selected 1-year-old painted maple and Manchurian walnut and 3-year-old Korean spruce according to the silvicultural practices of local forest farms. Younger broadleaved seedlings are used because of their relatively high growth rates compared with coniferous seedlings. However, painted maple is not often planted because it is not a commercial species. Seedling status before planting affected its final survival (Gerhardt, 1996). Seedlings consume stored carbohydrates to meet energy demands in stressful habitats, and a low storage could not support a long period of plant consumption (Myers and Kitajima, 2007). In our study, seedlings were planted before total leaf expansion (except Korean spruce, which is an evergreen species), and photosynthetic products may not satisfy energy consumption during the beginning period. By comparison, Sipe and Bazzaz (1995) collected and transplanted seedlings naturally regenerated from local forests, with age ranging from 4 to 20 years. The larger seedling size tended to have higher survival rates (Morrissey et al., 2010). According to their findings, the survival rates of shade-tolerant maples increased from forest understoreys to large gaps; within large gaps, seedling had higher survival rates in the southern part than in centre and northern part (Sipe and Bazzaz, 1995). The low survival rates in our study probably overrode the treatment effects and resulted in some random survival patterns within gaps. Thus, higher quality or larger painted maple seedlings should be planted in the future to detect whether the survival patterns within gaps are similar to other maple species. Seedling survival in gaps of different forest types Effects of gaps on seedling establishment were widely reported (Gray and Spies, 1996; Ouedraogo et al., 2014; Schwartz et al., 2017), but few of them compared the difference between forest types. We created gaps with similar size in larch plantations and secondary forests, and found that species had different survival dynamics in gaps of the same size but different forest types. For Manchurian walnut, seedlings had higher survival rates in larch plantation gaps than in secondary forest gaps during the study period, and the trend was similar to the comparison results between gaps of different sizes in secondary forests. The higher survival rates in larch plantations may be explained by the higher light intensity. The mean PAR in larch plantation gaps was 97.4 μM m−2 s−1 during two growing seasons, which was much higher than that in small secondary forest gaps (54.2 μM m−2 s−1) and even comparable with the mean PAR in medium (83.6 μM m−2 s−1) and large (114.9 μM m−2 s−1) secondary forest gaps. The mean DBH, stem density and basal area (>10 cm DBH) in secondary forests and larch plantations were 28 vs 18 cm, 351 vs 934 stem/ha and 24.4 vs 22.6 m2/ha. Although the larch plantation had higher stem density, it had similar basal area with the secondary forest. Thus, the higher light intensity in larch plantation understoreys was mainly due to the monopodial crown structure that permits more light to reach the ground (Bartemucci et al., 2002). Moreover, larch is an important timber species in Northeast China, and silvicultural treatments have been routinely prescribed to ensure plantation success (Mason and Zhu, 2014). For example, thinning was applied to reduce competition, promoting a higher growth rate for remaining crop trees (Chase et al., 2016). These silvicultural treatments also improved the light environment in larch plantation understoreys. A study on oak (Quercus ithaburensis Decne.) regeneration in pine (Pinus brutia Tenore) plantations found that improving the light environment for a few hours every day could significantly promote oak growth in the understoreys (Cooper et al., 2014a), and small gaps created by minimal pine removal were predicted to increase the stand structural complexity (Cooper et al., 2014b). In our study, the survival patterns in small gaps of two forest types were similar, with relatively lower survival rates in the west and south edges, which generally varied corresponding to the light distribution (Canham et al., 1990). These results indicated that forest types could affect seedling survival in gaps, but may not change the pattern of survival within gaps. The survival rates of Korean spruce were similar between gaps of two forest types during the study period, which were consistent with the comparison results in secondary forest gaps of different gap sizes and the results among within-gap positions. Combining the findings in secondary forest gaps, we inferred that the survival of Korean spruce could only benefit from gaps in a limited range and might not be significantly affected by other environmental differences between the two forest types. Although the survival rates of painted maple were higher in larch plantation gaps than in secondary forest gaps, and the survival patterns were different within gaps, significant statistical results based on such a low survival rate had no practical implications. The improved light environment in larch plantations did not increase its survival rate, which was consistent with the results in secondary forest gaps. These findings indicated that planting large and high-quality seedlings could be a better silvicultural practice in promoting seedling survival (South et al., 2005; Morrissey et al., 2010), after which forest types may be considered. Conclusions and management implications After eliminating the impact of vegetation competition and drought event, gaps could promote the survival of planted seedlings compared with forest understoreys, but the promoting effects differed among species, with the light-demanding species benefitting more than the two shade-tolerant species. Gap size, within-gap position and forest type mainly mattered to light-demanding Manchurian walnut. Shade-tolerant Korean spruce did not respond sensitively to positions within gaps and survived well in gaps of a wide size range and in gaps of different forest types. Shade-tolerant painted maple responded differently to gap size, within-gap position and forest type, but it had extremely low survival rates in all treatments. These conclusions have management implications for seedling establishment when a gap approach combined with enrichment planting and competition removal (e.g. weeding) is used. First, gap creation will benefit the survival of planted seedlings regardless of the target species. Second, gap size will only need to be considered if light-demanding species are planted. Third, planting light-demanding species in gap centres, transitions and north edge while shade-tolerant species in other gap edges could improve gap use efficiency and partially compensate smaller gap sizes. Finally, even for the same species, the optimum gap size may depend on stand conditions, especially the current level of light transmitted through the canopy. Smaller gaps may be created to promote seedling establishment in a well-managed stand (e.g. larch plantations in this study) where regular thinning has maintained small canopy openings and prevented the development of mid-storey. The current findings are based on a 2-year monitoring experiment. We will continue the study and evaluate whether this gap approach can be used for secondary forest restoration or larch plantation conversion in the future. Acknowledgements We thank Mr Chunyu Zhu and Mr Tao Yan from Qingyuan Forest CERN, Institute of Applied Ecology, Chinese Academy of Sciences for their help with the fieldwork. We thank Miss Brittany DiRienzo from Clemson University for the language editing of the article. We thank the editor and anonymous reviewers for their critical comments, especially the statistical analysis suggestions, on the article. Conflict of interest statement None declared. Funding National Natural Science Foundation of China (31 330 016), and Key Research Program of Frontier Sciences, CAS (QYZDJ-SSW-DQC027). References Adamic , M. , Diaci , J. , Rozman , A. and Hladnik , D. 2016 Long-term use of uneven-aged silviculture in mixed mountain Dinaric forests: a comparison of old-growth and managed stands . Forestry 90 ( 2 ), 279 – 291 . Arevalo , J.R. and Fernandez-Palacios , J.M. 2007 Treefall gaps and regeneration composition in the laurel forest of Anaga (Tenerife): a matter of size? Plant Ecol. 188 ( 2 ), 133 – 143 . Google Scholar CrossRef Search ADS Bartemucci , P. , Coates , K.D. , Harper , K.A. and Wright , E.F. 2002 Gap disturbances in northern old-growth forests of British Columbia, Canada . J. Veg. Sci. 13 ( 5 ), 685 – 696 . Google Scholar CrossRef Search ADS Bates , D. , Maechler , M. , Bolker , B. and Walker , S. 2015 Fitting linear mixed-effects models using lme4 . J. Stat. Softw. 67 ( 1 ), 1 – 48 . doi:10.18637/jss.v067.i01 . Google Scholar CrossRef Search ADS Bieling , C. , 2004 Non-industrial private-forest owners: possibilities for increasing adoption of close-to-nature forest management . Eur. J. For. Res. 123 ( 4 ), 293 – 303 . Google Scholar CrossRef Search ADS Bolton , N.W. and D’Amato , A.W. 2011 Regeneration responses to gap size and coarse woody debris within natural disturbance-based silvicultural systems in northeastern Minnesota, USA . For. Ecol. Manage. 262 ( 7 ), 1215 – 1222 . Google Scholar CrossRef Search ADS Brokaw , N.V.L. 1987 Gap-phase regeneration of 3 pioneer tree species in a tropical forest . J. Ecol. 75 ( 1 ), 9 – 19 . Google Scholar CrossRef Search ADS Brokaw , N. and Busing , R.T. 2000 Niche versus chance and tree diversity in forest gaps . Trends Ecol. Evol. 15 ( 5 ), 183 – 188 . Google Scholar CrossRef Search ADS PubMed Canham , C.D. , Denslow , J.S. , Platt , W.J. , Runkle , J.R. , Spies , T.A. and White , P.S. 1990 Light regimes beneath closed canopies and tree-fall gaps in temperate and tropical forests . Can. J. For. Res. 20 ( 5 ), 620 – 631 . Google Scholar CrossRef Search ADS Chase , C.W. , Kimsey , M.J. , Shaw , T.M. and Coleman , M.D. 2016 The response of light, water, and nutrient availability to pre-commercial thinning in dry inland Douglas-fir forests . For. Ecol. Manage. 363 , 98 – 109 . Google Scholar CrossRef Search ADS Chen , D.K. , Zhou , X.F. and Zhu , N. 1994 Natural Secondary Forest—Structure, Function, Dynamic and Management . Northeast Forestry University Press . (in Chinese). Cooper , A. , Moshe , Y. , Zangi , E. and Osem , Y. 2014 a Are small-scale overstory gaps effective in promoting the development of regenerating oaks (Quercus ithaburensis) in the forest understory? New Forests 45 ( 6 ), 843 – 857 . Google Scholar CrossRef Search ADS Cooper , A. , Shapira , O. , Zaidan , S. , Moshe , Y. , Zangy , E. and Osem , Y. 2014 b Oak restoration in water-limited pine plantations: interactive effects of overstory light interception and water availability on understory oak performance . Eur. J. For. Res. 133 ( 4 ), 661 – 670 . Google Scholar CrossRef Search ADS Curzon , M.T. and Keeton , W.S. 2010 Spatial characteristics of canopy disturbances in riparian old-growth hemlock—northern hardwood forests, Adirondack Mountains, New York, USA . Can. J. For. Res. 40 ( 1 ), 13 – 25 . Google Scholar CrossRef Search ADS Dauber , E. , Fredericksen , T.S. and Peña , M. 2005 Sustainability of timber harvesting in Bolivian tropical forests . For. Ecol. Manage. 214 ( 1–3 ), 294 – 304 . Google Scholar CrossRef Search ADS de Romer , A.H. , Kneeshaw , D.D. and Bergeron , Y. 2007 Small gap dynamics in the southern boreal forest of eastern Canada: do canopy gaps influence stand development? J. Veg. Sci. 18 ( 6 ), 815 – 826 . Google Scholar CrossRef Search ADS Denslow , J.S. 1980 Gap partitioning among tropical rainforest trees . Biotropica 12 , 47 – 55 . Google Scholar CrossRef Search ADS Diaci , J. , Gyoerek , N. , Gliha , J. and Nagel , T.A. 2008 Response of Quercus robur L. seedlings to north-south asymmetry of light within gaps in floodplain forests of Slovenia . Ann. For. Sci. 65 ( 1 ), 105 – 105 . Google Scholar CrossRef Search ADS Doucet , J.L. , Kouadio , Y.L. , Monticelli , D. and Lejeune , P. 2009 Enrichment of logging gaps with moabi (Baillonella toxisperma Pierre) in a Central African rain forest . For. Ecol. Manage. 258 ( 11 ), 2407 – 2415 . Google Scholar CrossRef Search ADS d’Oliveira , M.V.N. 2000 Artificial regeneration in gaps and skidding trails after mechanised forest exploitation in Acre, Brazil . For. Ecol. Manage. 127 ( 1–3 ), 67 – 76 . Google Scholar CrossRef Search ADS Galhidy , L. , Mihok , B. , Hagyo , A. , Rajkai , K. and Standovar , T. 2006 Effects of gap size and associated changes in light and soil moisture on the understorey vegetation of a Hungarian beech forest . Plant Ecol. 183 ( 1 ), 133 – 145 . Google Scholar CrossRef Search ADS Gerhardt , K. 1996 Effects of root competition and canopy openness on survival and growth of tree seedlings in a tropical seasonal dry forest . For. Ecol. Manage. 82 ( 1 ), 33 – 48 . Google Scholar CrossRef Search ADS Gray , A.N. and Spies , T.A. 1996 Gap size, within-gap position and canopy structure effects on conifer seedling establishment . J. Ecol. 84 ( 5 ), 635 – 645 . Google Scholar CrossRef Search ADS Heinemann , K. , Kitzberger , T. and Veblen , T.T. 2000 Influences of gap microheterogeneity on the regeneration of Nothofagus pumilio in a xeric old-growth forest of northwestern Patagonia, Argentina . Can. J. For. Res. 30 ( 1 ), 25 – 31 . Google Scholar CrossRef Search ADS Hubbell , S.P. , Foster , R.B. , O’Brien , S.T. , Harms , K.E. , Condit , R. , Wechsler , B. , et al. 1999 Light-gap disturbances, recruitment limitation, and tree diversity in a neotropical forest . Science 283 ( 5401 ), 554 – 557 . Google Scholar CrossRef Search ADS PubMed Kern , C.C. , Burton , J.I. , Raymond , P. , D’Amato , A.W. , Keeton , W.S. , Royo , A.A. , et al. 2016 Challenges facing gap-based silviculture and possible solutions for mesic northern forests in North America . Forestry 90 ( 1 ), 4 – 17 . Google Scholar CrossRef Search ADS Kern , C.C. , D’Arnato , A.W. and 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 . Google Scholar CrossRef Search ADS Kern , C.C. , Reich , P.B. , Montgomery , R.A. and Strong , T.F. 2012 Do deer and shrubs override canopy gap size effects on growth and survival of yellow birch, northern red oak, eastern white pine, and eastern hemlock seedlings? For. Ecol. Manage. 267 , 134 – 143 . Google Scholar CrossRef Search ADS Kitao , M. , Lei , T.T. , Koike , T. , Tobita , H. and Maruyama , Y. 2006 Tradeoff between shade adaptation and mitigation of photoinhibition in leaves of Quercus mongolica and Acer mono acclimated to deep shade . Tree Physiol. 26 ( 4 ), 441 – 448 . Google Scholar CrossRef Search ADS PubMed Knapp , B.O. , Wang , G.G. and 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 . Google Scholar CrossRef Search ADS Li , M.H. , Xiao , W.F. , Wang , S.G. , Cheng , G.W. , Cherubini , P. , Cai , X.H. , et al. 2008 Mobile carbohydrates in Himalayan treeline trees I. Evidence for carbon gain limitation but not for growth limitation . Tree Physiol. 28 ( 8 ), 1287 – 1296 . Google Scholar CrossRef Search ADS PubMed Liu , C. , Zhu , J. , Wu , X. and Yang , K. 2011 Characteristics of soil nutrients within canopy gaps of various sizes in secondary forests in eastern mountainous regions of Liaoning Province, China . J. Northeast For. Univ. 39 ( 1 ), 79 – 81 . Lobo , E. and Dalling , J.W. 2013 Effects of topography, soil type and forest age on the frequency and size distribution of canopy gap disturbances in a tropical forest . Biogeosciences 10 ( 11 ), 6769 – 6781 . Google Scholar CrossRef Search ADS Madsen , P. and Hahn , K. 2008 Natural regeneration in a beech-dominated forest managed by close-to-nature principles—a gap cutting based experiment . Can. J. For. Res 38 ( 7 ), 1716 – 1729 . Google Scholar CrossRef Search ADS Mason , W.L. and 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 . Fenning T. (ed). Springer , pp. 37 – 54 . Google Scholar CrossRef Search ADS McCarthy , J. 2001 Gap dynamics of forest trees: a review with particular attention to boreal forests . Environ. Rev. 9 ( 1 ), 1 – 59 . Google Scholar CrossRef Search ADS Morrissey , R.C. , Jacobs , D.F. , Davis , A.S. and Rathfon , R.A. 2010 Survival and competitiveness of Quercus rubra regeneration associated with planting stocktype and harvest opening intensity . New Forests 40 ( 3 ), 273 – 287 . Google Scholar CrossRef Search ADS Muscolo , A. , Sidari , M. and Mercurio , R. 2007 Influence of gap size on organic matter decomposition, microbial biomass and nutrient cycle in Calabrian pine (Pinus laricio, Poiret) stands . For. Ecol. Manage. 242 ( 2 ), 412 – 418 . Google Scholar CrossRef Search ADS Myers , J.A. and Kitajima , K. 2007 Carbohydrate storage enhances seedling shade and stress tolerance in a neotropical forest . J. Ecol. 95 ( 2 ), 383 – 395 . Google Scholar CrossRef Search ADS Obrien , M.J. , Leuzinger , S. , Philipson , C.D. , Tay , J. and Hector , A. 2014 Drought survival of tropical tree seedlings enhanced by non-structural carbohydrate levels . Nat. Clim. Change 4 ( 8 ), 710 – 714 . Google Scholar CrossRef Search ADS Ouedraogo , D.Y. , Fayolle , A. , Dainou , K. , Demaret , C. , Bourland , N. , Lagoute , P. , et al. 2014 Enrichment of logging gaps with a high conservation value species (Pericopsis elata) in a central African moist forest . Forests 5 ( 12 ), 3031 – 3047 . Google Scholar CrossRef Search ADS Paquette , A. , Bouchard , A. and Cogliastro , A. 2006 Survival and growth of under-planted trees: a meta-analysis across four biomes . Ecol. Appl. 16 ( 4 ), 1575 – 1589 . Google Scholar CrossRef Search ADS PubMed R Core Team 2016 R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing . Vienna , https://www.R-project.org/. Runkle , J.R. and Yetter , T.C. 1987 Treefalls revisited: gap dynamics in the Southern Appalachians . Ecology 68 ( 2 ), 417 – 424 . Google Scholar CrossRef Search ADS Schnitzer , S.A. and Carson , W.P. 2001 Treefall gaps and the maintenance of species diversity in a tropical forest . Ecology 82 ( 4 ), 913 – 919 . Google Scholar CrossRef Search ADS 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 ( 3–4 ), 866 – 879 . Google Scholar CrossRef Search ADS Schwartz , G. , Pereira , P.C.G. , Siviero , M.A. , Pereira , J.F. , Ruschel , A.R. and Yared , J.A.G. 2017 Enrichment planting in logging gaps with Schizolobium parahyba var. amazonicum (Huber ex Ducke) Barneby: a financially profitable alternative for degraded tropical forests in the Amazon . For. Ecol. Manage. 390 , 166 – 172 . Google Scholar CrossRef Search ADS Sharma , L.N. , Grytnes , J.-A. , Måren , I.E. and Vetaas , O.R. 2016 Do composition and richness of woody plants vary between gaps and closed canopy patches in subtropical forests? J. Veg. Sci. 27 ( 6 ), 1129 – 1139 . Google Scholar CrossRef Search ADS Sipe , T. and Bazzaz , F. 1995 Gap partitioning among maples (Acer) in central New England: survival and growth . Ecology 76 ( 5 ), 1587 – 1602 . Google Scholar CrossRef Search ADS South , D.B. , Harris , S.W. , Barnett , J.P. , Hainds , M.J. and Gjerstad , D.H. 2005 Effect of container type and seedling size on survival and early height growth of Pinus palustris seedlings in Alabama, U.S.A . For. Ecol. Manage. 204 ( 2–3 ), 385 – 398 . Google Scholar CrossRef Search ADS Streit , K. , Wunder , J. and Brang , P. 2009 Slit-shaped gaps are a successful silvicultural technique to promote Picea abies regeneration in mountain forests of the Swiss Alps . For. Ecol. Manage. 257 ( 9 ), 1902 – 1909 . Google Scholar CrossRef Search ADS Therneau , T.M. 2018 coxme: Mixed Effects Cox Models. R package version 2.2–7. https://CRAN.R-project.org/package=coxme Vepakomma , U. , Kneeshaw , D. and Fortin , M.J. 2012 Spatial contiguity and continuity of canopy gaps in mixed wood boreal forests: persistence, expansion, shrinkage and displacement . J. Ecol 100 ( 5 ), 1257 – 1268 . Google Scholar CrossRef Search ADS Vilhar , U. , Roženbergar , D. , Simončič , P. and Diaci , J. 2015 Variation in irradiance, soil features and regeneration patterns in experimental forest canopy gaps . Ann. Forest Sci. 72 ( 2 ), 253 – 266 . Google Scholar CrossRef Search ADS Wang , G.L. and Liu , F. 2011 The influence of gap creation on the regeneration of Pinus tabuliformis planted forest and its role in the near-natural cultivation strategy for planted forest management . For. Ecol. Manage. 262 ( 3 ), 413 – 423 . Google Scholar CrossRef Search ADS Weber , T.A. , Hart , J.L. , Schweitzer , C.J. and Dey , D.C. 2014 Influence of gap-scale disturbance on developmental and successional pathways in Quercus-Pinus stands . For. Ecol. Manage. 331 , 60 – 70 . Google Scholar CrossRef Search ADS Webster , C.R. and Lorimer , C.G. 2005 Minimum opening sizes for canopy recruitment of midtolerant tree species: a retrospective approach . Ecol. Appl. 15 ( 4 ), 1245 – 1262 . Google Scholar CrossRef Search ADS Wright , E.F. , Coates , K.D. and Bartemucci , P. 1998 Regeneration from seed of six tree species in the interior cedar-hemlock forests of British Columbia as affected by substrate and canopy gap position . Can. J. For. Res. 28 ( 9 ), 1352 – 1364 . Google Scholar CrossRef Search ADS York , R.A. , Battles , J.J. , Eschtruth , A.K. and Schurr , F.G. 2011 Giant sequoia (Sequoiadendron giganteum) regeneration in experimental canopy gaps . Restor. Ecol. 19 ( 1 ), 14 – 23 . Google Scholar CrossRef Search ADS Zhu , J.J. , Liu , Z.G. , Wang , H.X. , Yan , Q.L. , Fang , H.Y. , Hu , L.L. , et al. 2008 Effects of site preparation on emergence and early establishment of Larix olgensis in montane regions of northeastern China . New Forests 36 , 247 – 260 . Google Scholar CrossRef Search ADS Zhu , J.J. , Lu , D.L. and Zhang , W.D. 2014 a Effects of gaps on regeneration of woody plants: a meta-analysis . J. Forestry Res. 25 ( 3 ), 501 – 510 . Google Scholar CrossRef Search ADS Zhu , J.J. , Wang , K. , Sun , Y.R. and Yan , Q.L. 2014 b Response of Pinus koraiensis seedling growth to different light conditions based on the assessment of photosynthesis in current and one-year-old needles . J. Forestry Res. 25 ( 1 ), 53 – 62 . Google Scholar CrossRef Search ADS © Institute of Chartered Foresters, 2018. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Forestry: An International Journal Of Forest Research Oxford University Press

Seedling survival within forest gaps: the effects of gap size, within-gap position and forest type on species of contrasting shade-tolerance in Northeast China

Loading next page...
 
/lp/ou_press/seedling-survival-within-forest-gaps-the-effects-of-gap-size-within-tL7rUxYuL7
Publisher
Institute of Chartered Foresters
Copyright
© Institute of Chartered Foresters, 2018. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.
ISSN
0015-752X
eISSN
1464-3626
D.O.I.
10.1093/forestry/cpy007
Publisher site
See Article on Publisher Site

Abstract

Abstract Previous studies on gap regeneration have seldom addressed potential differences due to within-gap position, condition of forest type and within-gap vegetation competition. We excluded understorey vegetation competition, and tested the effects of gap size, within-gap position and forest type on the survival of three species with contrasting shade tolerance by planting seedlings in twelve gaps (four each of small, medium and large size) created in secondary forests and four small gaps created in larch (Larix spp.) plantations. In the absence of vegetation competition, the survival of planted seedlings was generally higher in gaps than in forest understoreys despite there being significant differences among species. Light-demanding Juglans mandshurica survived better in large and medium gaps than in small gaps, and in the gap centre, transition, north edge and east edge than in other within-gap positions. Moreover, seedlings in larch plantation gaps had higher survival rates than in secondary forest gaps of the same size. Shade-tolerant Picea koraiensis had high survival rates that were not affected by gap size, within-gap position and forest type. Shade-tolerant Acer mono had extremely low survival rates in all treatments. Our findings suggest that, in addition to gap size, within-gap position and forest type condition should be considered to promote seedling establishment of light-demanding species during enrichment planting within gaps. Introduction Forest gaps, often resulting from a small-scale disturbance, are commonly found in forests throughout the world (Brokaw, 1987; Runkle and Yetter, 1987; McCarthy, 2001). Ecological effects of forest gaps have been extensively studied in tropical (Brokaw, 1987; Schnitzer and Carson, 2001), temperate (Runkle and Yetter, 1987; Weber et al., 2014) and boreal (de Romer et al., 2007) forests. Generally, gaps increased seedling or sapling density or diversity and helped maintain the stability of forest ecosystems (Adamic et al., 2016), although insignificant or even negative gap effects on tree regeneration were also reported (Hubbell et al., 1999; Arevalo and Fernandez-Palacios, 2007). However, the distributions of natural gaps are complex (Curzon and Keeton, 2010; Lobo and Dalling, 2013), and gap characteristics, such as size, age and shape, vary greatly (Vepakomma et al., 2012). In addition, the conditions of the stand where gaps are located vary considerably (Lobo and Dalling, 2013). All these uncertainties may limit the generality of findings from gap studies and constrain the application of findings in forest management. As the concept of close-to-nature silviculture develops, emulating natural disturbances becomes popular in modern forest management. One approach is to create artificial gaps based on stand conditions to achieve forest regeneration and other management objectives (Bieling, 2004; Streit et al., 2009). For example, slit-shaped gaps as a silvicultural technique were successful in increasing the seedling density of Picea abies (L.) Karst. in the Swiss Alps (Streit et al., 2009). In the northeast US, yellow birch (Betula alleghaniensis Britt.) was able to grow to the canopy in hemlock (Tsuga canadensis L.) gaps created by group selection (Webster and Lorimer, 2005). However, natural regeneration in gaps may depend on chance events (Brokaw and Busing, 2000), and desired tree species rarely appear without additional silvicultural interventions (Dauber et al., 2005). For example, a gap creation experiment did not increase the density of target tree species after 5 years (Madsen and Hahn, 2008), as many factors (e.g. lacking seed resource) may potentially weaken or invalidate the gap effects on natural regeneration. Enrichment planting (i.e. planting trees in forest stand gaps to increase the percentage of desired species or biodiversity) can increase the establishment of target species in gaps, which can be especially important for degraded forests (d’Oliveira, 2000; Ouedraogo et al., 2014; Schwartz et al., 2017). Seeds sowed or seedlings planted within gaps had higher survival rates and diameter growth increments after a few years (Schulze, 2008; Doucet et al., 2009; Ouedraogo et al., 2014). However, most of these studies focused on gap centre areas, with the exception of Gray and Spies (1996) and York et al. (2011), which may overestimate the gap effects on enrichment planting. The microenvironment within a gap is not homogeneous, especially in northern latitude forests (Canham et al., 1990; Brokaw and Busing, 2000; Diaci et al., 2008). Therefore, growth conditions at any given place within a gap would depend on the gap size and within-gap position. Furthermore, seedling survival and growth of desired species are often limited by competing vegetation within forest gaps, and the development of the competing vegetation within a gap also depends on the gap size and within-gap position. Evaluating gap centre effects or the whole gap effects would not reveal seedling performance among gaps of different sizes or different within-gap positions. For example, in the northern hemisphere, seedling survival may be higher in the southern part of gaps, where light intensity was modest but soil moisture was high (Gray and Spies, 1996; Wright et al., 1998). Although the gap partitioning hypothesis suggests that shade-tolerant species survive better in small gaps or forest understoreys compared with large gaps (Denslow, 1980), results from previous studies proved to be highly inconsistent (Wright et al., 1998; Zhu et al., 2014a; Sharma et al., 2016). We select secondary forests as our study object. Secondary forests are the largest ecosystem type in Northeast China, which function as important water conservation forests as well as timber reserves. Historically the main silvicultural approach focused on maximizing timber production, but it has been gradually changed to providing ecological service through close-to-nature silviculture such as gap creation. In this study, we eliminate vegetation competition and evaluate the effects of gap size, within-gap position, and forest type on the survival of three planted seedling species with different shade tolerance. Specifically, we address the following questions: (1) Can gaps increase the survival of planted seedlings compared with forest understoreys? (2) How do gap size, within-gap position and forest type affect seedling survival? (3) How does shade tolerance affect seedling survival? We hypothesize that gaps could promote seedling survival regardless of shade tolerance; light-demanding species may survive the highest in large gaps and gap centres, but shade-tolerant species would not show significant survival difference among gap sizes or within-gap positions. We expect that gap creation combined with enrichment planting and competition removal could be an appropriate approach for seedling establishment in the early stage. Methods Study site The experiment was carried out in Qingyuan Forest, the Chinese Ecosystem Research Network (CERN) site established by the Chinese Academy of Sciences (CAS). The study area is located within a mountain region in Liaoning Province, Northeast China (41°51′ N, 124°54′ E), with elevation ranging from 500 to 1100 m above sea level. It has a temperate monsoon climate, characterized by a warm and humid summer, and a cold and dry winter. The mean annual air temperature is 4.7°C. The coldest month is January, with a mean air temperature of −12.1°C, and the warmest month is July, with a mean air temperature of 21.0°C. The annual precipitation is 811 mm, and 80 per cent of precipitation falls during the summer. The frost-free period is 130 days, with an early frost occurring in October and a late frost occurring in April. The study area was characterized by mixed broadleaved-Korean pine (Pinus koraiensis Sieb. et Zucc.) forests, which were the largest natural forest community in Northeast China (Chen et al., 1994; Zhu et al., 2008). However, more than 70 per cent of these forests were severely degraded with almost no remaining natural Korean pine trees after a century’s timber exploitation (Chen et al., 1994; Zhu et al., 2008). Deciduous natural secondary forests and non-contiguous patches of larch (Larix spp.) plantations are currently dominating the study area. The dominant species in the secondary forests include ash (Fraxinus rhynchophylla Hance), Manchurian walnut (Juglans mandshurica Maxim.), Mongolian oak (Quercus mongolica Fisch.) and painted maple (Acer mono Maxim.) (Zhu et al., 2014b). The main plantation species are larch, Korean pine and Korean spruce (Picea koraiensis Nakai). Forests in this area are mainly disturbed by winds. Snowstorms and floods occur infrequently. Forest fires are rare. Disease, insect and browse are not common issues for the local species. Gap creation Overall, 16 gaps were artificially created in March 2015 (Table 1): 12 gaps in secondary forest stands with three size classes (large gaps: 658–984 m2, medium gaps: 476–512 m2 and small gaps: 176–260 m2), and four small gaps in larch plantation stands (130–206 m2). The mean ratio of gap diameter to gap border tree height (RD/H) of large, medium and small gaps in secondary forests is 1.66, 1.36 and 0.92, respectively. The mean RD/H of small gaps in larch plantations is 0.71. No large or medium gaps were created in larch plantations because a cutting area >200 m2 is prohibited before harvesting by the local Forestry Administration Agency. All gap locations were randomly selected in the forests with generally similar topography, slope and aspect. During the gap creation, all trees and shrubs taller than 30 cm within the gaps were cut off and removed. Every gap border tree was tagged, and its species, height and diameter at breast height were recorded. To minimize the negative effects of gap creation on forest soil and other trees, we cut down trees with chainsaws and moved them out of the gaps manually instead of using machines. The expanded gap size and shape were measured with a total station (TKS-202, China) as follows: (1) select an estimated gap centre point; (2) measure the direction and distance between the gap centre point and one gap border tree, and scale the data to coordinate paper; (3) repeat the previous procedure until all gap border trees are measured; and (4) delineate the gap shape, calculate the gap size and adjust the gap centre point. With the help of a total station, gaps could be created into any shape, and we approximated each gap consistently as a circle. Table 1 General description of the forest gaps created in March 2015 aForest gap Expanded gap size (m2) Altitude (m) Slope (°) Aspect (°) Mean height of gap border trees (m) Mean DBH of gap border trees (cm) bSpecies composition of gap border trees (%) L 984 678 22 191 20 27.3 QM (47), JM (24), FR (18) L 968 653 19 195 18 28.9 QM (77), FR (9), PD (9) L 688 679 25 170 20 30.8 QM (53), JM (47) L 658 672 25 220 20 32.6 QM (100) M 512 731 28 158 17 26.5 UP (35), JM (24), QM (18) M 492 762 26 170 19 32.5 QM (63), JM (19), FR (13) M 484 726 25 170 19 31.6 QM (47), JM (27), UP (20) M 476 664 18 168 19 28.8 QM (50), FM (25), BP (19) S 260 699 28 228 20 27.1 QM (77), LS (23) S 226 686 28 195 17 20.3 JM (63), FR (19), FM (13) S 184 708 22 208 17 21.1 FM (58), QM (25), FR (17) S 176 713 27 165 17 25.8 QM (75), FR (17), UP (8) S (Larch) 206 719 25 190 20 17 LS (100) S (Larch) 156 669 25 164 20 18.6 LS (100) S (Larch) 142 734 15 190 20 18.6 LS (100) S (Larch) 130 698 25 183 20 17.4 LS (100) aForest gap Expanded gap size (m2) Altitude (m) Slope (°) Aspect (°) Mean height of gap border trees (m) Mean DBH of gap border trees (cm) bSpecies composition of gap border trees (%) L 984 678 22 191 20 27.3 QM (47), JM (24), FR (18) L 968 653 19 195 18 28.9 QM (77), FR (9), PD (9) L 688 679 25 170 20 30.8 QM (53), JM (47) L 658 672 25 220 20 32.6 QM (100) M 512 731 28 158 17 26.5 UP (35), JM (24), QM (18) M 492 762 26 170 19 32.5 QM (63), JM (19), FR (13) M 484 726 25 170 19 31.6 QM (47), JM (27), UP (20) M 476 664 18 168 19 28.8 QM (50), FM (25), BP (19) S 260 699 28 228 20 27.1 QM (77), LS (23) S 226 686 28 195 17 20.3 JM (63), FR (19), FM (13) S 184 708 22 208 17 21.1 FM (58), QM (25), FR (17) S 176 713 27 165 17 25.8 QM (75), FR (17), UP (8) S (Larch) 206 719 25 190 20 17 LS (100) S (Larch) 156 669 25 164 20 18.6 LS (100) S (Larch) 142 734 15 190 20 18.6 LS (100) S (Larch) 130 698 25 183 20 17.4 LS (100) aL = large gap; M = medium gap; S = small gap; S (Larch) = small gap in larch plantations. bBP = Betula platyphylla Suk.; FM = Fraxinus mandshurica Rupr.; FR = Fraxinus rhynchophylla Hance; JM = Juglans mandshurica Maxim.; LS = Larix spp.; PD = Populus davidiana Dode; QM = Quercus mongolica Fisch. ex Ledeb.; UP = Ulmus pumila L. Table 1 General description of the forest gaps created in March 2015 aForest gap Expanded gap size (m2) Altitude (m) Slope (°) Aspect (°) Mean height of gap border trees (m) Mean DBH of gap border trees (cm) bSpecies composition of gap border trees (%) L 984 678 22 191 20 27.3 QM (47), JM (24), FR (18) L 968 653 19 195 18 28.9 QM (77), FR (9), PD (9) L 688 679 25 170 20 30.8 QM (53), JM (47) L 658 672 25 220 20 32.6 QM (100) M 512 731 28 158 17 26.5 UP (35), JM (24), QM (18) M 492 762 26 170 19 32.5 QM (63), JM (19), FR (13) M 484 726 25 170 19 31.6 QM (47), JM (27), UP (20) M 476 664 18 168 19 28.8 QM (50), FM (25), BP (19) S 260 699 28 228 20 27.1 QM (77), LS (23) S 226 686 28 195 17 20.3 JM (63), FR (19), FM (13) S 184 708 22 208 17 21.1 FM (58), QM (25), FR (17) S 176 713 27 165 17 25.8 QM (75), FR (17), UP (8) S (Larch) 206 719 25 190 20 17 LS (100) S (Larch) 156 669 25 164 20 18.6 LS (100) S (Larch) 142 734 15 190 20 18.6 LS (100) S (Larch) 130 698 25 183 20 17.4 LS (100) aForest gap Expanded gap size (m2) Altitude (m) Slope (°) Aspect (°) Mean height of gap border trees (m) Mean DBH of gap border trees (cm) bSpecies composition of gap border trees (%) L 984 678 22 191 20 27.3 QM (47), JM (24), FR (18) L 968 653 19 195 18 28.9 QM (77), FR (9), PD (9) L 688 679 25 170 20 30.8 QM (53), JM (47) L 658 672 25 220 20 32.6 QM (100) M 512 731 28 158 17 26.5 UP (35), JM (24), QM (18) M 492 762 26 170 19 32.5 QM (63), JM (19), FR (13) M 484 726 25 170 19 31.6 QM (47), JM (27), UP (20) M 476 664 18 168 19 28.8 QM (50), FM (25), BP (19) S 260 699 28 228 20 27.1 QM (77), LS (23) S 226 686 28 195 17 20.3 JM (63), FR (19), FM (13) S 184 708 22 208 17 21.1 FM (58), QM (25), FR (17) S 176 713 27 165 17 25.8 QM (75), FR (17), UP (8) S (Larch) 206 719 25 190 20 17 LS (100) S (Larch) 156 669 25 164 20 18.6 LS (100) S (Larch) 142 734 15 190 20 18.6 LS (100) S (Larch) 130 698 25 183 20 17.4 LS (100) aL = large gap; M = medium gap; S = small gap; S (Larch) = small gap in larch plantations. bBP = Betula platyphylla Suk.; FM = Fraxinus mandshurica Rupr.; FR = Fraxinus rhynchophylla Hance; JM = Juglans mandshurica Maxim.; LS = Larix spp.; PD = Populus davidiana Dode; QM = Quercus mongolica Fisch. ex Ledeb.; UP = Ulmus pumila L. Plot set up and seedling planting Each gap was divided into three areas with nine positions identified (Figure 1): gap centre (one position), gap transition (four positions: north, south, east and west), and gap edge (four positions: north, south, east and west). All positions were determined by their direction and relative distance from the gap centre. If the distance between gap centre point and each gap border tree was 1.0, the coordinates of 0.0, 0.5 and 1.0 on the north–centre–south line and east–centre–west line were the positions for gap centre, gap transition and gap edge, respectively. Figure 1 View largeDownload slide Schematic diagram of plot distribution within each gap. Figure 1 View largeDownload slide Schematic diagram of plot distribution within each gap. In each gap position, one 3 m × 3 m plot with a 0.2-m-wide buffer zone was set up for seedling planting. Eight plots were set up in the forest understoreys of secondary forests and larch plantations (four for each) as control. Totally, 152 plots were established in our experiment. All control plots were placed under the canopy without disturbance and more than 20 m distance from gaps, which assured the plot environment was not affected by gaps. Litter and woody debris within the plots resulted from the operation of gap creation were cleaned manually before planting. Competitive species (mainly herbs) in all plots including forest understoreys were removed monthly during two growing seasons after planting, because our previous experiment indicated that planted seedlings could be severely suppressed by herbs during the early stage, which could potentially negate the effect of gap size or within-gap position. Seedlings of three major species in Northeast China, painted maple, Manchurian walnut, and Korean spruce, were used for enrichment planting during late April and early May of 2015. Each plot was divided into six columns including 36 squares (0.5 m × 0.5 m), and each column was assigned one species randomly according to a random number sequence. For Manchurian walnut and Korean spruce, one seedling was planted in each square (12 seedlings per plot). For painted maple, two additional seedlings were planted in each square (36 seedlings per plot). As a result, a total of 5472 painted maple, 1824 Manchurian walnut and 1824 Korean spruce seedlings were planted. The number of painted maple seedlings was tripled because our early experiment found a relatively low survival rate of planted painted maple seedlings. Immediately after planting, seedlings of all plots were watered thoroughly because of abnormally dry soil conditions at the time of planting, although watering was not included in our original experimental design. All seedlings were bare-rooted seedlings, and each species was transplanted from the same local nursery. Painted maple and Manchurian walnut were 1-year-old seedlings, and Korean spruce were 3-year-old seedlings. During the 2-year study period, seedlings planted within each plot had maintained enough spacing to avoid competing with each other for light. Seedling measurement before planting In total, 12 seedlings of each species were selected and divided into three groups for measuring the seedling parameters before planting (Table 2). Height, basal diameter, biomass and non-structural carbohydrate (NSC) pool were measured. The NSC pool (in milligrams) was calculated as seedling biomass × NSC concentration, which was determined by anthrone colourimetry method (Li et al., 2008). Table 2 Seedling status before planting Species name Shade tolerance Age (y) aMean basal diameter (mm) bMean height (cm) cMean biomass (g) dMean NSC pool (mg) Acer mono Shade tolerant 1 3.31 (0.15) 21.80 (0.90) 1.69 (0.04) 46.98 (10.90) Juglans mandshurica Light demanding 1 6.32 (0.40) 31.00 (1.53) 5.85 (0.69) 439.24 (30.48) Picea koraiensis Shade tolerant 3 6.78 (0.61) 20.33 (0.88) 13.47 (0.33) 1121.21 (38.01) Species name Shade tolerance Age (y) aMean basal diameter (mm) bMean height (cm) cMean biomass (g) dMean NSC pool (mg) Acer mono Shade tolerant 1 3.31 (0.15) 21.80 (0.90) 1.69 (0.04) 46.98 (10.90) Juglans mandshurica Light demanding 1 6.32 (0.40) 31.00 (1.53) 5.85 (0.69) 439.24 (30.48) Picea koraiensis Shade tolerant 3 6.78 (0.61) 20.33 (0.88) 13.47 (0.33) 1121.21 (38.01) a,b,c,dValues in parenthesis represent standard error (N = 3). Table 2 Seedling status before planting Species name Shade tolerance Age (y) aMean basal diameter (mm) bMean height (cm) cMean biomass (g) dMean NSC pool (mg) Acer mono Shade tolerant 1 3.31 (0.15) 21.80 (0.90) 1.69 (0.04) 46.98 (10.90) Juglans mandshurica Light demanding 1 6.32 (0.40) 31.00 (1.53) 5.85 (0.69) 439.24 (30.48) Picea koraiensis Shade tolerant 3 6.78 (0.61) 20.33 (0.88) 13.47 (0.33) 1121.21 (38.01) Species name Shade tolerance Age (y) aMean basal diameter (mm) bMean height (cm) cMean biomass (g) dMean NSC pool (mg) Acer mono Shade tolerant 1 3.31 (0.15) 21.80 (0.90) 1.69 (0.04) 46.98 (10.90) Juglans mandshurica Light demanding 1 6.32 (0.40) 31.00 (1.53) 5.85 (0.69) 439.24 (30.48) Picea koraiensis Shade tolerant 3 6.78 (0.61) 20.33 (0.88) 13.47 (0.33) 1121.21 (38.01) a,b,c,dValues in parenthesis represent standard error (N = 3). Seedling survival and light environment monitoring Seedling survival was examined monthly for two growing seasons (2015 and 2016). If no green tissue on the seedling stem was observable (some seedlings were re-observed for 1 or 2 months), the seedling was regarded as dead (Obrien et al., 2014). The seedling survival rate of each species in each plot was calculated as follows: Seedlingsurvivalrate(%)=NtN0×100% (1) where, Nt is the number of seedlings that survived during the investigation at time, t, and N0 is the number of seedling initially planted. One gap of each size class and one control of each forest type were randomly selected for environmental monitoring. A data logger (WatchDog 1650 Micro Station, Spectrum Technologies, Inc., USA) was set up in each plot to continuously monitor photosynthetically active radiation (PAR) at 1 m above the ground, soil temperature at 5 cm below the ground, and soil moisture at 5 cm below the ground (Table 3). Table 3 Mean light, soil temperature and soil moisture environment in gaps and forest understoreys during two growing seasons Position Photosynthetically active radiation (PAR, μM m−2 s−1) Soil temperature at 5 cm (°C) Soil moisture at 5 cm (%) aLarge bMedium cSmall dSmall (Larch) Large Medium Small Small (Larch) Large Medium Small Small (Larch) Centre 174.53 141.39 90.31 123.83 18.67 18.44 18.38 17.82 29.75 27.09 17.43 20.79 North transition 179.71 125.79 94.31 155.87 19.45 18.38 18.82 18.58 19.13 12.44 16.60 25.92 East transition 150.43 108.50 75.90 93.02 19.57 18.98 17.85 17.61 15.37 15.19 18.81 30.47 West transition 148.45 127.68 62.76 121.61 18.40 17.68 18.07 18.11 32.17 21.98 20.97 19.36 South transition 119.74 91.61 41.65 69.84 17.81 16.99 17.41 17.04 28.18 41.48 11.80 24.07 North edge 67.18 43.59 28.51 102.96 18.33 17.39 16.46 18.89 13.76 15.40 13.50 9.18 East edge 84.86 28.06 41.82 61.08 18.63 17.51 17.51 17.55 9.04 17.19 19.63 12.00 West edge 54.75 50.03 31.01 94.97 17.63 17.19 17.10 17.91 13.91 14.23 10.70 9.90 South edge 54.73 35.96 21.10 53.43 17.45 16.78 16.83 17.10 25.43 35.99 15.68 11.23 eForest understorey 13.10 28.80 16.96 16.98 14.42 13.86 Position Photosynthetically active radiation (PAR, μM m−2 s−1) Soil temperature at 5 cm (°C) Soil moisture at 5 cm (%) aLarge bMedium cSmall dSmall (Larch) Large Medium Small Small (Larch) Large Medium Small Small (Larch) Centre 174.53 141.39 90.31 123.83 18.67 18.44 18.38 17.82 29.75 27.09 17.43 20.79 North transition 179.71 125.79 94.31 155.87 19.45 18.38 18.82 18.58 19.13 12.44 16.60 25.92 East transition 150.43 108.50 75.90 93.02 19.57 18.98 17.85 17.61 15.37 15.19 18.81 30.47 West transition 148.45 127.68 62.76 121.61 18.40 17.68 18.07 18.11 32.17 21.98 20.97 19.36 South transition 119.74 91.61 41.65 69.84 17.81 16.99 17.41 17.04 28.18 41.48 11.80 24.07 North edge 67.18 43.59 28.51 102.96 18.33 17.39 16.46 18.89 13.76 15.40 13.50 9.18 East edge 84.86 28.06 41.82 61.08 18.63 17.51 17.51 17.55 9.04 17.19 19.63 12.00 West edge 54.75 50.03 31.01 94.97 17.63 17.19 17.10 17.91 13.91 14.23 10.70 9.90 South edge 54.73 35.96 21.10 53.43 17.45 16.78 16.83 17.10 25.43 35.99 15.68 11.23 eForest understorey 13.10 28.80 16.96 16.98 14.42 13.86 a,b,cLarge, medium and small represent large gaps, medium gaps and small gaps in secondary forests, respectively. dSmall (Larch) represents small gaps in larch plantations. eForest understorey represents the control plots of each forest type. Table 3 Mean light, soil temperature and soil moisture environment in gaps and forest understoreys during two growing seasons Position Photosynthetically active radiation (PAR, μM m−2 s−1) Soil temperature at 5 cm (°C) Soil moisture at 5 cm (%) aLarge bMedium cSmall dSmall (Larch) Large Medium Small Small (Larch) Large Medium Small Small (Larch) Centre 174.53 141.39 90.31 123.83 18.67 18.44 18.38 17.82 29.75 27.09 17.43 20.79 North transition 179.71 125.79 94.31 155.87 19.45 18.38 18.82 18.58 19.13 12.44 16.60 25.92 East transition 150.43 108.50 75.90 93.02 19.57 18.98 17.85 17.61 15.37 15.19 18.81 30.47 West transition 148.45 127.68 62.76 121.61 18.40 17.68 18.07 18.11 32.17 21.98 20.97 19.36 South transition 119.74 91.61 41.65 69.84 17.81 16.99 17.41 17.04 28.18 41.48 11.80 24.07 North edge 67.18 43.59 28.51 102.96 18.33 17.39 16.46 18.89 13.76 15.40 13.50 9.18 East edge 84.86 28.06 41.82 61.08 18.63 17.51 17.51 17.55 9.04 17.19 19.63 12.00 West edge 54.75 50.03 31.01 94.97 17.63 17.19 17.10 17.91 13.91 14.23 10.70 9.90 South edge 54.73 35.96 21.10 53.43 17.45 16.78 16.83 17.10 25.43 35.99 15.68 11.23 eForest understorey 13.10 28.80 16.96 16.98 14.42 13.86 Position Photosynthetically active radiation (PAR, μM m−2 s−1) Soil temperature at 5 cm (°C) Soil moisture at 5 cm (%) aLarge bMedium cSmall dSmall (Larch) Large Medium Small Small (Larch) Large Medium Small Small (Larch) Centre 174.53 141.39 90.31 123.83 18.67 18.44 18.38 17.82 29.75 27.09 17.43 20.79 North transition 179.71 125.79 94.31 155.87 19.45 18.38 18.82 18.58 19.13 12.44 16.60 25.92 East transition 150.43 108.50 75.90 93.02 19.57 18.98 17.85 17.61 15.37 15.19 18.81 30.47 West transition 148.45 127.68 62.76 121.61 18.40 17.68 18.07 18.11 32.17 21.98 20.97 19.36 South transition 119.74 91.61 41.65 69.84 17.81 16.99 17.41 17.04 28.18 41.48 11.80 24.07 North edge 67.18 43.59 28.51 102.96 18.33 17.39 16.46 18.89 13.76 15.40 13.50 9.18 East edge 84.86 28.06 41.82 61.08 18.63 17.51 17.51 17.55 9.04 17.19 19.63 12.00 West edge 54.75 50.03 31.01 94.97 17.63 17.19 17.10 17.91 13.91 14.23 10.70 9.90 South edge 54.73 35.96 21.10 53.43 17.45 16.78 16.83 17.10 25.43 35.99 15.68 11.23 eForest understorey 13.10 28.80 16.96 16.98 14.42 13.86 a,b,cLarge, medium and small represent large gaps, medium gaps and small gaps in secondary forests, respectively. dSmall (Larch) represents small gaps in larch plantations. eForest understorey represents the control plots of each forest type. Data analysis Generalized linear mixed-effect models (logistic regressions with binomial distribution and logit link function) were used to compare the final seedling survival of each species between gaps and forest understoreys in secondary forests and larch plantations. Survival analysis models (Mixed-effect Cox regressions based on condition independence through time) were used for small gaps in two forest types to examine the effects of forest type and species on seedling survival. Similarly, survival analysis models were used for all secondary forest gaps to examine the effects of gap size and species on seedling survival. Finally, Logistic regressions were used to compare the final seedling survival of each species between different within-gap positions. For all the analyses, gap was regarded as a random effect. A P < 0.05 was regarded as statistically significant. All the analyses were performed with R version 3.3.2 (R Core Team, 2016). The coxme package (Therneau, 2018) was used for survival analyses. The lme4 package (Bates et al., 2015) was used for logistic regressions. Results Seedling survival in gaps and forest understoreys Logistic regressions indicated that the final seedling survival rates were significantly higher in gaps than forest understoreys in both secondary forests and larch plantations for Korean spruce (P < 0.05, Table 4) and Manchurian walnut (P < 0.001, Table 4). The final seedling survival rates of painted maple were significantly higher in secondary forest gaps than understoreys (P < 0.01, Table 4), but there was only a trend of significant survival difference between larch plantation gaps and understoreys (P = 0.058, Table 4). The mean final survival rates in gaps/understoreys were 61.4 per cent/45.0 per cent, 48.2 per cent/17.5 per cent and 8.0 per cent/1.4 per cent in secondary forests, and 62.8 per cent/42.5 per cent, 52.2 per cent/20.0 per cent and 9.8 per cent/4.9 per cent in larch plantations of Korean spruce, Manchurian walnut and painted maple, respectively. Table 4 Seedling survival estimates of logistic regressions for three planted species in larch plantations (4 gaps and 4 controls) and secondary forests (12 gaps and 4 controls). Significant effects (P < 0.05) are in bold Species Model term Larch plantation Secondary forest Coefficient SE Wald Z P-value Coefficient SE Wald Z P-value Picea koraiensis Understorey (intercept) −0.33 0.33 −0.98 0.329 −0.20 0.33 −0.62 0.537 Gap 0.85 0.36 2.38 0.018 0.67 0.33 2.03 0.042 Juglans mandshurica Understorey (intercept) −1.39 0.39 −3.51 <0.001 −1.81 0.45 −3.99 <0.001 Gap 1.48 0.41 3.61 <0.001 1.74 0.44 3.97 <0.001 Acer mono Understorey (intercept) −2.97 0.39 −7.67 <0.001 −4.36 0.72 −6.08 <0.001 Gap 0.75 0.40 1.89 0.058 1.89 0.72 2.64 0.008 Species Model term Larch plantation Secondary forest Coefficient SE Wald Z P-value Coefficient SE Wald Z P-value Picea koraiensis Understorey (intercept) −0.33 0.33 −0.98 0.329 −0.20 0.33 −0.62 0.537 Gap 0.85 0.36 2.38 0.018 0.67 0.33 2.03 0.042 Juglans mandshurica Understorey (intercept) −1.39 0.39 −3.51 <0.001 −1.81 0.45 −3.99 <0.001 Gap 1.48 0.41 3.61 <0.001 1.74 0.44 3.97 <0.001 Acer mono Understorey (intercept) −2.97 0.39 −7.67 <0.001 −4.36 0.72 −6.08 <0.001 Gap 0.75 0.40 1.89 0.058 1.89 0.72 2.64 0.008 Table 4 Seedling survival estimates of logistic regressions for three planted species in larch plantations (4 gaps and 4 controls) and secondary forests (12 gaps and 4 controls). Significant effects (P < 0.05) are in bold Species Model term Larch plantation Secondary forest Coefficient SE Wald Z P-value Coefficient SE Wald Z P-value Picea koraiensis Understorey (intercept) −0.33 0.33 −0.98 0.329 −0.20 0.33 −0.62 0.537 Gap 0.85 0.36 2.38 0.018 0.67 0.33 2.03 0.042 Juglans mandshurica Understorey (intercept) −1.39 0.39 −3.51 <0.001 −1.81 0.45 −3.99 <0.001 Gap 1.48 0.41 3.61 <0.001 1.74 0.44 3.97 <0.001 Acer mono Understorey (intercept) −2.97 0.39 −7.67 <0.001 −4.36 0.72 −6.08 <0.001 Gap 0.75 0.40 1.89 0.058 1.89 0.72 2.64 0.008 Species Model term Larch plantation Secondary forest Coefficient SE Wald Z P-value Coefficient SE Wald Z P-value Picea koraiensis Understorey (intercept) −0.33 0.33 −0.98 0.329 −0.20 0.33 −0.62 0.537 Gap 0.85 0.36 2.38 0.018 0.67 0.33 2.03 0.042 Juglans mandshurica Understorey (intercept) −1.39 0.39 −3.51 <0.001 −1.81 0.45 −3.99 <0.001 Gap 1.48 0.41 3.61 <0.001 1.74 0.44 3.97 <0.001 Acer mono Understorey (intercept) −2.97 0.39 −7.67 <0.001 −4.36 0.72 −6.08 <0.001 Gap 0.75 0.40 1.89 0.058 1.89 0.72 2.64 0.008 Seedling survival in secondary forest gaps of different sizes Cox regressions indicated that seedling survival was significantly affected by gap size and species during the study period (P < 0.001). Compared with small gaps, the mean hazard of mortality reduced 15 and 11 per cent in large and medium gaps, respectively (Table 5). Compared with painted maple, the mean hazard of mortality reduced 65 and 80 per cent for Manchurian walnut and Korean spruce, respectively (Table 5). Manchurian walnut had significantly higher survival rate in large and medium gaps than in small gaps (P < 0.001; Figure 2). No significant survival difference was found for Korean spruce between large, medium and small gaps (P > 0.60; Figure 2). Painted maple had significantly higher survival rate in large and medium gaps than in small gaps (P < 0.05; Figure 2). The rank of survival rates among species did not change during the study period (Figure 2). Korean spruce had the highest survival rate, and painted maple had the lowest survival rate with remarkable mortality occurring 30 days after planting (Figure 2). Table 5 Survival results of Cox regressions for three planted species in gaps of different sizes and forest types. The analysis of small gaps in two forest types includes 4320 seedlings with 943 seedlings surviving beyond the last census. The analysis of gaps of different sizes in secondary forests includes 6480 seedlings with 1493 seedlings surviving beyond the last census. Seedling survival was assessed monthly during each growing season (2015 and 2016). Significant effects (P < 0.05) are in bold Model term Coefficient dEXP (coefficient) SE (coefficient) Wald Z P-value Gaps of different sizes in secondary forests  aSize – medium −0.12 0.89 0.06 −2.14 0.032  Size – large −0.16 0.85 0.06 −2.99 0.003  bSpecies – JM −1.05 0.35 0.06 −16.59 <0.001  Species – PK −1.62 0.20 0.07 −22.16 <0.001  Size – medium: species – JM −0.32 0.73 0.10 −3.36 <0.001  Size – large: species – JM −0.39 0.68 0.10 −4.01 <0.001  Size – medium: species – PK −0.05 0.95 0.11 −0.47 0.630  Size – large: species – PK 0.04 1.05 0.11 0.43 0.670 Small gaps in two forest types  cForest type – Larch −0.14 0.87 0.05 −2.72 0.007  Species – JM −1.06 0.35 0.06 −16.63 <0.001  Species – PK −1.62 0.20 0.07 −22.15 <0.001  Forest type – Larch: species – JM −0.26 0.77 0.09 −2.80 0.005  Forest type – Larch: species – PK 0.01 1.01 0.10 0.08 0.940 Model term Coefficient dEXP (coefficient) SE (coefficient) Wald Z P-value Gaps of different sizes in secondary forests  aSize – medium −0.12 0.89 0.06 −2.14 0.032  Size – large −0.16 0.85 0.06 −2.99 0.003  bSpecies – JM −1.05 0.35 0.06 −16.59 <0.001  Species – PK −1.62 0.20 0.07 −22.16 <0.001  Size – medium: species – JM −0.32 0.73 0.10 −3.36 <0.001  Size – large: species – JM −0.39 0.68 0.10 −4.01 <0.001  Size – medium: species – PK −0.05 0.95 0.11 −0.47 0.630  Size – large: species – PK 0.04 1.05 0.11 0.43 0.670 Small gaps in two forest types  cForest type – Larch −0.14 0.87 0.05 −2.72 0.007  Species – JM −1.06 0.35 0.06 −16.63 <0.001  Species – PK −1.62 0.20 0.07 −22.15 <0.001  Forest type – Larch: species – JM −0.26 0.77 0.09 −2.80 0.005  Forest type – Larch: species – PK 0.01 1.01 0.10 0.08 0.940 a,b,cGap size – small, species – AM and forest type – secondary forest are baselines. dEXP (coefficient) indicates the mean hazard of mortality relative to baselines. Table 5 Survival results of Cox regressions for three planted species in gaps of different sizes and forest types. The analysis of small gaps in two forest types includes 4320 seedlings with 943 seedlings surviving beyond the last census. The analysis of gaps of different sizes in secondary forests includes 6480 seedlings with 1493 seedlings surviving beyond the last census. Seedling survival was assessed monthly during each growing season (2015 and 2016). Significant effects (P < 0.05) are in bold Model term Coefficient dEXP (coefficient) SE (coefficient) Wald Z P-value Gaps of different sizes in secondary forests  aSize – medium −0.12 0.89 0.06 −2.14 0.032  Size – large −0.16 0.85 0.06 −2.99 0.003  bSpecies – JM −1.05 0.35 0.06 −16.59 <0.001  Species – PK −1.62 0.20 0.07 −22.16 <0.001  Size – medium: species – JM −0.32 0.73 0.10 −3.36 <0.001  Size – large: species – JM −0.39 0.68 0.10 −4.01 <0.001  Size – medium: species – PK −0.05 0.95 0.11 −0.47 0.630  Size – large: species – PK 0.04 1.05 0.11 0.43 0.670 Small gaps in two forest types  cForest type – Larch −0.14 0.87 0.05 −2.72 0.007  Species – JM −1.06 0.35 0.06 −16.63 <0.001  Species – PK −1.62 0.20 0.07 −22.15 <0.001  Forest type – Larch: species – JM −0.26 0.77 0.09 −2.80 0.005  Forest type – Larch: species – PK 0.01 1.01 0.10 0.08 0.940 Model term Coefficient dEXP (coefficient) SE (coefficient) Wald Z P-value Gaps of different sizes in secondary forests  aSize – medium −0.12 0.89 0.06 −2.14 0.032  Size – large −0.16 0.85 0.06 −2.99 0.003  bSpecies – JM −1.05 0.35 0.06 −16.59 <0.001  Species – PK −1.62 0.20 0.07 −22.16 <0.001  Size – medium: species – JM −0.32 0.73 0.10 −3.36 <0.001  Size – large: species – JM −0.39 0.68 0.10 −4.01 <0.001  Size – medium: species – PK −0.05 0.95 0.11 −0.47 0.630  Size – large: species – PK 0.04 1.05 0.11 0.43 0.670 Small gaps in two forest types  cForest type – Larch −0.14 0.87 0.05 −2.72 0.007  Species – JM −1.06 0.35 0.06 −16.63 <0.001  Species – PK −1.62 0.20 0.07 −22.15 <0.001  Forest type – Larch: species – JM −0.26 0.77 0.09 −2.80 0.005  Forest type – Larch: species – PK 0.01 1.01 0.10 0.08 0.940 a,b,cGap size – small, species – AM and forest type – secondary forest are baselines. dEXP (coefficient) indicates the mean hazard of mortality relative to baselines. Figure 2 View largeDownload slide Survival rates in secondary forest gaps of different sizes (N = 4) for Acer mono (AM), Juglans mandshurica (JM) and Picea koraiensis (PK). Figure 2 View largeDownload slide Survival rates in secondary forest gaps of different sizes (N = 4) for Acer mono (AM), Juglans mandshurica (JM) and Picea koraiensis (PK). Logistic regressions indicated that Manchurian walnut finally had lower survival rates in west edge and south edge, but the survival rates in other within-gap positions were not significantly different from the gap centre (Figure 3). For Manchurian walnut, the survival patterns were similar in large, medium and small gaps (Figure 3). For Korean spruce, no significant survival differences were found among within-gap positions in large, medium and small gaps (Figure 3). The survival of painted maple within gaps showed random patterns. For example, seedling had lower survival rates in the west transition of small gaps, in the south edge and east edge of medium gaps, but no significant survival differences were found within large gaps (Figure 3). Figure 3 View largeDownload slide Final survival rates of nine within-gap positions (N = 4) for all gaps in larch plantations and secondary forests. Significant effects (P < 0.05; compared with the survival rate in gap centre which is regarded as the baseline) are in bold. Figure 3 View largeDownload slide Final survival rates of nine within-gap positions (N = 4) for all gaps in larch plantations and secondary forests. Significant effects (P < 0.05; compared with the survival rate in gap centre which is regarded as the baseline) are in bold. Seedling survival in small gaps of two forest types Cox regressions indicated that seedling survival was significantly affected by forest type and species during the study period (P < 0.001). Compared with secondary forests, the mean hazard of mortality reduced 13 per cent in larch plantations (Table 5). Compared with painted maple, the mean hazard of mortality reduced 65 and 80 per cent for Manchurian walnut and Korean spruce, respectively (Table 5). Manchurian walnut had significantly higher survival rates in larch plantation gaps than in secondary forest gaps (P < 0.01; Figure 4). No significant survival difference was found for Korean spruce between gaps in secondary forests and larch plantations (P > 0.90; Figure 4). Painted maple had significantly higher survival rates in larch plantation gaps than in secondary forest gaps (P < 0.01; Figure 4). The rank of seedling survival among species was similar in gaps of two forest types (Figure 4). Figure 4 View largeDownload slide Survival rates of small gaps in larch plantations (Lar; N = 4) and secondary forests (Sec; N = 4) for Acer mono (AM), Juglans mandshurica (JM) and Picea koraiensis (PK). Figure 4 View largeDownload slide Survival rates of small gaps in larch plantations (Lar; N = 4) and secondary forests (Sec; N = 4) for Acer mono (AM), Juglans mandshurica (JM) and Picea koraiensis (PK). Logistic regressions indicated that the final survival rates of Manchurian walnut were significantly lower in the west edge, south edge and east edge of larch plantation gaps, but the survival rates in other within-gap positions were not significantly different from the gap centre (Figure 3). The final survival patterns of Manchurian walnut were similar in two forest types (Figure 3). For Korean spruce, no significant survival differences were found among within-gap positions in larch plantation gaps, which was similar to the patterns in small secondary forest gaps (Figure 3). There was no significant survival difference among within-gap positions for painted maple in larch plantation gaps, which differed from small secondary forest gaps (Figure 3). Discussion Seedling survival in gaps and forest understoreys Enrichment planting in artificial gaps is an efficient way to promote seedling establishment in an existing stand. Gaps improve the microenvironment which affects nutrient release and seedling establishment (Muscolo et al., 2007; Liu et al., 2011; Vilhar et al., 2015). In our study, almost all species survived better in gaps than in forest understoreys regardless of shade tolerance. These findings supported that the regeneration of shade-tolerant species could also benefit from gaps (Wang and Liu, 2011). Although all studied species had positive responses to gaps compared with forest understoreys, the increase of seedling survival rates of the two shade-tolerant species (painted maple and Korean spruce) were lower than the light-demanding species (Manchurian walnut). These results were consistent with a recent meta-analysis, which reported that the relative density of light-demanding species in gaps was significantly higher than shade-tolerant species (Zhu et al., 2014a). Seedling survival in relation to gap sizes and within-gap positions Gap size is one of the most important factors affecting seedling regeneration within gaps (Bolton and D’Amato, 2011; Kern et al., 2012, 2016). However, previous studies indicated that the effect of gap size on regeneration was confounded by several factors (Bolton and D’Amato, 2011; Kern et al., 2016). For example, creating large gaps may reduce tree density due to the increased resource competition from other vegetation, such as shrubs and herbs (Kern et al., 2013). Therefore, we excluded the competition effect by manually removing non-target species from planting plots to further examine seedling responses to gaps. In addition, the environment within gaps is spatially heterogeneous, especially at high latitudes (Canham et al., 1990; Heinemann et al., 2000). Our study was carried out in the Northern Hemisphere, and all gaps had general south aspects, which may further amplify the environmental differences among within-gap positions, especially along the north–centre–south line within gaps. Therefore, it was necessary to examine the effects of within-gap positions on seedling survival. For light-demanding Manchurian walnut, the survival rate decreased faster in small gaps compared with large and medium gaps over the 2-year study period. Seedlings planted in the south edge and west edge usually had lower survival rates than those in other within-gap positions at the end of the second growing season, which was generally consistent with the light distribution within gaps (Canham et al., 1990; Diaci et al., 2008). These results indicated that the survival of light-demanding species benefited from the higher light supply in larger gaps or within-gap positions with more exposure (Galhidy et al., 2006; Cooper et al., 2014a, 2014b). However, Knapp et al. (2013) found that the mortality of planted Pinus palustris Mill. seedlings was higher in the gap centre than gap edge, and in the northern part compared with the southern part of gaps, but they attributed their results to climatic drought during the study period. A global meta-analysis indicated that annual seedling survival rates could differ significantly throughout different biomes even under the same canopy treatment (Paquette et al., 2006). Therefore, the effects of gap size and within-gap position on seedling establishment must be interpreted with the context and limitation of each particular study. Although shade-tolerant Korean spruce had higher survival rates in gaps than in forest understoreys, similar survival rates were found in gaps of different sizes and in different within-gap positions over the 2 years. A previous study focusing on regeneration of shade-tolerant coniferous species within gaps also found higher seedling survival rates in gaps than in forest understoreys, but all the species were more abundant in the shaded parts within gaps (Gray and Spies, 1996). These findings indicated that shade-tolerant coniferous species, such as Korean spruce, could benefit from increased light intensity only within a limited range, but may not partition along gap resource gradient (e.g. light intensity) as shown in light-demanding species. Previous studies reported that shade-tolerant painted maple could survive in forest understoreys due to high photosynthetic rate at low light environment (Kitao et al., 2006). However, the mean final survival rates of painted maple were generally <10 per cent in our study. Although significant survival differences existed between forest types, gap sizes and within-gap positions, the extremely low survival rates of painted maple had little meaning in practice. A previous study monitored the survival of three maple species (Acer pensylvanicum, Acer rubrum and Acer saccharum) planted in gaps and reported a mean survival rate of 79 per cent after three years (Sipe and Bazzaz, 1995), which is much higher than our results. The low survival rates of painted maple may be caused by the much smaller seedling dimensions, biomass and NSC pool at the time of planting. We selected 1-year-old painted maple and Manchurian walnut and 3-year-old Korean spruce according to the silvicultural practices of local forest farms. Younger broadleaved seedlings are used because of their relatively high growth rates compared with coniferous seedlings. However, painted maple is not often planted because it is not a commercial species. Seedling status before planting affected its final survival (Gerhardt, 1996). Seedlings consume stored carbohydrates to meet energy demands in stressful habitats, and a low storage could not support a long period of plant consumption (Myers and Kitajima, 2007). In our study, seedlings were planted before total leaf expansion (except Korean spruce, which is an evergreen species), and photosynthetic products may not satisfy energy consumption during the beginning period. By comparison, Sipe and Bazzaz (1995) collected and transplanted seedlings naturally regenerated from local forests, with age ranging from 4 to 20 years. The larger seedling size tended to have higher survival rates (Morrissey et al., 2010). According to their findings, the survival rates of shade-tolerant maples increased from forest understoreys to large gaps; within large gaps, seedling had higher survival rates in the southern part than in centre and northern part (Sipe and Bazzaz, 1995). The low survival rates in our study probably overrode the treatment effects and resulted in some random survival patterns within gaps. Thus, higher quality or larger painted maple seedlings should be planted in the future to detect whether the survival patterns within gaps are similar to other maple species. Seedling survival in gaps of different forest types Effects of gaps on seedling establishment were widely reported (Gray and Spies, 1996; Ouedraogo et al., 2014; Schwartz et al., 2017), but few of them compared the difference between forest types. We created gaps with similar size in larch plantations and secondary forests, and found that species had different survival dynamics in gaps of the same size but different forest types. For Manchurian walnut, seedlings had higher survival rates in larch plantation gaps than in secondary forest gaps during the study period, and the trend was similar to the comparison results between gaps of different sizes in secondary forests. The higher survival rates in larch plantations may be explained by the higher light intensity. The mean PAR in larch plantation gaps was 97.4 μM m−2 s−1 during two growing seasons, which was much higher than that in small secondary forest gaps (54.2 μM m−2 s−1) and even comparable with the mean PAR in medium (83.6 μM m−2 s−1) and large (114.9 μM m−2 s−1) secondary forest gaps. The mean DBH, stem density and basal area (>10 cm DBH) in secondary forests and larch plantations were 28 vs 18 cm, 351 vs 934 stem/ha and 24.4 vs 22.6 m2/ha. Although the larch plantation had higher stem density, it had similar basal area with the secondary forest. Thus, the higher light intensity in larch plantation understoreys was mainly due to the monopodial crown structure that permits more light to reach the ground (Bartemucci et al., 2002). Moreover, larch is an important timber species in Northeast China, and silvicultural treatments have been routinely prescribed to ensure plantation success (Mason and Zhu, 2014). For example, thinning was applied to reduce competition, promoting a higher growth rate for remaining crop trees (Chase et al., 2016). These silvicultural treatments also improved the light environment in larch plantation understoreys. A study on oak (Quercus ithaburensis Decne.) regeneration in pine (Pinus brutia Tenore) plantations found that improving the light environment for a few hours every day could significantly promote oak growth in the understoreys (Cooper et al., 2014a), and small gaps created by minimal pine removal were predicted to increase the stand structural complexity (Cooper et al., 2014b). In our study, the survival patterns in small gaps of two forest types were similar, with relatively lower survival rates in the west and south edges, which generally varied corresponding to the light distribution (Canham et al., 1990). These results indicated that forest types could affect seedling survival in gaps, but may not change the pattern of survival within gaps. The survival rates of Korean spruce were similar between gaps of two forest types during the study period, which were consistent with the comparison results in secondary forest gaps of different gap sizes and the results among within-gap positions. Combining the findings in secondary forest gaps, we inferred that the survival of Korean spruce could only benefit from gaps in a limited range and might not be significantly affected by other environmental differences between the two forest types. Although the survival rates of painted maple were higher in larch plantation gaps than in secondary forest gaps, and the survival patterns were different within gaps, significant statistical results based on such a low survival rate had no practical implications. The improved light environment in larch plantations did not increase its survival rate, which was consistent with the results in secondary forest gaps. These findings indicated that planting large and high-quality seedlings could be a better silvicultural practice in promoting seedling survival (South et al., 2005; Morrissey et al., 2010), after which forest types may be considered. Conclusions and management implications After eliminating the impact of vegetation competition and drought event, gaps could promote the survival of planted seedlings compared with forest understoreys, but the promoting effects differed among species, with the light-demanding species benefitting more than the two shade-tolerant species. Gap size, within-gap position and forest type mainly mattered to light-demanding Manchurian walnut. Shade-tolerant Korean spruce did not respond sensitively to positions within gaps and survived well in gaps of a wide size range and in gaps of different forest types. Shade-tolerant painted maple responded differently to gap size, within-gap position and forest type, but it had extremely low survival rates in all treatments. These conclusions have management implications for seedling establishment when a gap approach combined with enrichment planting and competition removal (e.g. weeding) is used. First, gap creation will benefit the survival of planted seedlings regardless of the target species. Second, gap size will only need to be considered if light-demanding species are planted. Third, planting light-demanding species in gap centres, transitions and north edge while shade-tolerant species in other gap edges could improve gap use efficiency and partially compensate smaller gap sizes. Finally, even for the same species, the optimum gap size may depend on stand conditions, especially the current level of light transmitted through the canopy. Smaller gaps may be created to promote seedling establishment in a well-managed stand (e.g. larch plantations in this study) where regular thinning has maintained small canopy openings and prevented the development of mid-storey. The current findings are based on a 2-year monitoring experiment. We will continue the study and evaluate whether this gap approach can be used for secondary forest restoration or larch plantation conversion in the future. Acknowledgements We thank Mr Chunyu Zhu and Mr Tao Yan from Qingyuan Forest CERN, Institute of Applied Ecology, Chinese Academy of Sciences for their help with the fieldwork. We thank Miss Brittany DiRienzo from Clemson University for the language editing of the article. We thank the editor and anonymous reviewers for their critical comments, especially the statistical analysis suggestions, on the article. Conflict of interest statement None declared. Funding National Natural Science Foundation of China (31 330 016), and Key Research Program of Frontier Sciences, CAS (QYZDJ-SSW-DQC027). References Adamic , M. , Diaci , J. , Rozman , A. and Hladnik , D. 2016 Long-term use of uneven-aged silviculture in mixed mountain Dinaric forests: a comparison of old-growth and managed stands . Forestry 90 ( 2 ), 279 – 291 . Arevalo , J.R. and Fernandez-Palacios , J.M. 2007 Treefall gaps and regeneration composition in the laurel forest of Anaga (Tenerife): a matter of size? Plant Ecol. 188 ( 2 ), 133 – 143 . Google Scholar CrossRef Search ADS Bartemucci , P. , Coates , K.D. , Harper , K.A. and Wright , E.F. 2002 Gap disturbances in northern old-growth forests of British Columbia, Canada . J. Veg. Sci. 13 ( 5 ), 685 – 696 . Google Scholar CrossRef Search ADS Bates , D. , Maechler , M. , Bolker , B. and Walker , S. 2015 Fitting linear mixed-effects models using lme4 . J. Stat. Softw. 67 ( 1 ), 1 – 48 . doi:10.18637/jss.v067.i01 . Google Scholar CrossRef Search ADS Bieling , C. , 2004 Non-industrial private-forest owners: possibilities for increasing adoption of close-to-nature forest management . Eur. J. For. Res. 123 ( 4 ), 293 – 303 . Google Scholar CrossRef Search ADS Bolton , N.W. and D’Amato , A.W. 2011 Regeneration responses to gap size and coarse woody debris within natural disturbance-based silvicultural systems in northeastern Minnesota, USA . For. Ecol. Manage. 262 ( 7 ), 1215 – 1222 . Google Scholar CrossRef Search ADS Brokaw , N.V.L. 1987 Gap-phase regeneration of 3 pioneer tree species in a tropical forest . J. Ecol. 75 ( 1 ), 9 – 19 . Google Scholar CrossRef Search ADS Brokaw , N. and Busing , R.T. 2000 Niche versus chance and tree diversity in forest gaps . Trends Ecol. Evol. 15 ( 5 ), 183 – 188 . Google Scholar CrossRef Search ADS PubMed Canham , C.D. , Denslow , J.S. , Platt , W.J. , Runkle , J.R. , Spies , T.A. and White , P.S. 1990 Light regimes beneath closed canopies and tree-fall gaps in temperate and tropical forests . Can. J. For. Res. 20 ( 5 ), 620 – 631 . Google Scholar CrossRef Search ADS Chase , C.W. , Kimsey , M.J. , Shaw , T.M. and Coleman , M.D. 2016 The response of light, water, and nutrient availability to pre-commercial thinning in dry inland Douglas-fir forests . For. Ecol. Manage. 363 , 98 – 109 . Google Scholar CrossRef Search ADS Chen , D.K. , Zhou , X.F. and Zhu , N. 1994 Natural Secondary Forest—Structure, Function, Dynamic and Management . Northeast Forestry University Press . (in Chinese). Cooper , A. , Moshe , Y. , Zangi , E. and Osem , Y. 2014 a Are small-scale overstory gaps effective in promoting the development of regenerating oaks (Quercus ithaburensis) in the forest understory? New Forests 45 ( 6 ), 843 – 857 . Google Scholar CrossRef Search ADS Cooper , A. , Shapira , O. , Zaidan , S. , Moshe , Y. , Zangy , E. and Osem , Y. 2014 b Oak restoration in water-limited pine plantations: interactive effects of overstory light interception and water availability on understory oak performance . Eur. J. For. Res. 133 ( 4 ), 661 – 670 . Google Scholar CrossRef Search ADS Curzon , M.T. and Keeton , W.S. 2010 Spatial characteristics of canopy disturbances in riparian old-growth hemlock—northern hardwood forests, Adirondack Mountains, New York, USA . Can. J. For. Res. 40 ( 1 ), 13 – 25 . Google Scholar CrossRef Search ADS Dauber , E. , Fredericksen , T.S. and Peña , M. 2005 Sustainability of timber harvesting in Bolivian tropical forests . For. Ecol. Manage. 214 ( 1–3 ), 294 – 304 . Google Scholar CrossRef Search ADS de Romer , A.H. , Kneeshaw , D.D. and Bergeron , Y. 2007 Small gap dynamics in the southern boreal forest of eastern Canada: do canopy gaps influence stand development? J. Veg. Sci. 18 ( 6 ), 815 – 826 . Google Scholar CrossRef Search ADS Denslow , J.S. 1980 Gap partitioning among tropical rainforest trees . Biotropica 12 , 47 – 55 . Google Scholar CrossRef Search ADS Diaci , J. , Gyoerek , N. , Gliha , J. and Nagel , T.A. 2008 Response of Quercus robur L. seedlings to north-south asymmetry of light within gaps in floodplain forests of Slovenia . Ann. For. Sci. 65 ( 1 ), 105 – 105 . Google Scholar CrossRef Search ADS Doucet , J.L. , Kouadio , Y.L. , Monticelli , D. and Lejeune , P. 2009 Enrichment of logging gaps with moabi (Baillonella toxisperma Pierre) in a Central African rain forest . For. Ecol. Manage. 258 ( 11 ), 2407 – 2415 . Google Scholar CrossRef Search ADS d’Oliveira , M.V.N. 2000 Artificial regeneration in gaps and skidding trails after mechanised forest exploitation in Acre, Brazil . For. Ecol. Manage. 127 ( 1–3 ), 67 – 76 . Google Scholar CrossRef Search ADS Galhidy , L. , Mihok , B. , Hagyo , A. , Rajkai , K. and Standovar , T. 2006 Effects of gap size and associated changes in light and soil moisture on the understorey vegetation of a Hungarian beech forest . Plant Ecol. 183 ( 1 ), 133 – 145 . Google Scholar CrossRef Search ADS Gerhardt , K. 1996 Effects of root competition and canopy openness on survival and growth of tree seedlings in a tropical seasonal dry forest . For. Ecol. Manage. 82 ( 1 ), 33 – 48 . Google Scholar CrossRef Search ADS Gray , A.N. and Spies , T.A. 1996 Gap size, within-gap position and canopy structure effects on conifer seedling establishment . J. Ecol. 84 ( 5 ), 635 – 645 . Google Scholar CrossRef Search ADS Heinemann , K. , Kitzberger , T. and Veblen , T.T. 2000 Influences of gap microheterogeneity on the regeneration of Nothofagus pumilio in a xeric old-growth forest of northwestern Patagonia, Argentina . Can. J. For. Res. 30 ( 1 ), 25 – 31 . Google Scholar CrossRef Search ADS Hubbell , S.P. , Foster , R.B. , O’Brien , S.T. , Harms , K.E. , Condit , R. , Wechsler , B. , et al. 1999 Light-gap disturbances, recruitment limitation, and tree diversity in a neotropical forest . Science 283 ( 5401 ), 554 – 557 . Google Scholar CrossRef Search ADS PubMed Kern , C.C. , Burton , J.I. , Raymond , P. , D’Amato , A.W. , Keeton , W.S. , Royo , A.A. , et al. 2016 Challenges facing gap-based silviculture and possible solutions for mesic northern forests in North America . Forestry 90 ( 1 ), 4 – 17 . Google Scholar CrossRef Search ADS Kern , C.C. , D’Arnato , A.W. and 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 . Google Scholar CrossRef Search ADS Kern , C.C. , Reich , P.B. , Montgomery , R.A. and Strong , T.F. 2012 Do deer and shrubs override canopy gap size effects on growth and survival of yellow birch, northern red oak, eastern white pine, and eastern hemlock seedlings? For. Ecol. Manage. 267 , 134 – 143 . Google Scholar CrossRef Search ADS Kitao , M. , Lei , T.T. , Koike , T. , Tobita , H. and Maruyama , Y. 2006 Tradeoff between shade adaptation and mitigation of photoinhibition in leaves of Quercus mongolica and Acer mono acclimated to deep shade . Tree Physiol. 26 ( 4 ), 441 – 448 . Google Scholar CrossRef Search ADS PubMed Knapp , B.O. , Wang , G.G. and 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 . Google Scholar CrossRef Search ADS Li , M.H. , Xiao , W.F. , Wang , S.G. , Cheng , G.W. , Cherubini , P. , Cai , X.H. , et al. 2008 Mobile carbohydrates in Himalayan treeline trees I. Evidence for carbon gain limitation but not for growth limitation . Tree Physiol. 28 ( 8 ), 1287 – 1296 . Google Scholar CrossRef Search ADS PubMed Liu , C. , Zhu , J. , Wu , X. and Yang , K. 2011 Characteristics of soil nutrients within canopy gaps of various sizes in secondary forests in eastern mountainous regions of Liaoning Province, China . J. Northeast For. Univ. 39 ( 1 ), 79 – 81 . Lobo , E. and Dalling , J.W. 2013 Effects of topography, soil type and forest age on the frequency and size distribution of canopy gap disturbances in a tropical forest . Biogeosciences 10 ( 11 ), 6769 – 6781 . Google Scholar CrossRef Search ADS Madsen , P. and Hahn , K. 2008 Natural regeneration in a beech-dominated forest managed by close-to-nature principles—a gap cutting based experiment . Can. J. For. Res 38 ( 7 ), 1716 – 1729 . Google Scholar CrossRef Search ADS Mason , W.L. and 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 . Fenning T. (ed). Springer , pp. 37 – 54 . Google Scholar CrossRef Search ADS McCarthy , J. 2001 Gap dynamics of forest trees: a review with particular attention to boreal forests . Environ. Rev. 9 ( 1 ), 1 – 59 . Google Scholar CrossRef Search ADS Morrissey , R.C. , Jacobs , D.F. , Davis , A.S. and Rathfon , R.A. 2010 Survival and competitiveness of Quercus rubra regeneration associated with planting stocktype and harvest opening intensity . New Forests 40 ( 3 ), 273 – 287 . Google Scholar CrossRef Search ADS Muscolo , A. , Sidari , M. and Mercurio , R. 2007 Influence of gap size on organic matter decomposition, microbial biomass and nutrient cycle in Calabrian pine (Pinus laricio, Poiret) stands . For. Ecol. Manage. 242 ( 2 ), 412 – 418 . Google Scholar CrossRef Search ADS Myers , J.A. and Kitajima , K. 2007 Carbohydrate storage enhances seedling shade and stress tolerance in a neotropical forest . J. Ecol. 95 ( 2 ), 383 – 395 . Google Scholar CrossRef Search ADS Obrien , M.J. , Leuzinger , S. , Philipson , C.D. , Tay , J. and Hector , A. 2014 Drought survival of tropical tree seedlings enhanced by non-structural carbohydrate levels . Nat. Clim. Change 4 ( 8 ), 710 – 714 . Google Scholar CrossRef Search ADS Ouedraogo , D.Y. , Fayolle , A. , Dainou , K. , Demaret , C. , Bourland , N. , Lagoute , P. , et al. 2014 Enrichment of logging gaps with a high conservation value species (Pericopsis elata) in a central African moist forest . Forests 5 ( 12 ), 3031 – 3047 . Google Scholar CrossRef Search ADS Paquette , A. , Bouchard , A. and Cogliastro , A. 2006 Survival and growth of under-planted trees: a meta-analysis across four biomes . Ecol. Appl. 16 ( 4 ), 1575 – 1589 . Google Scholar CrossRef Search ADS PubMed R Core Team 2016 R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing . Vienna , https://www.R-project.org/. Runkle , J.R. and Yetter , T.C. 1987 Treefalls revisited: gap dynamics in the Southern Appalachians . Ecology 68 ( 2 ), 417 – 424 . Google Scholar CrossRef Search ADS Schnitzer , S.A. and Carson , W.P. 2001 Treefall gaps and the maintenance of species diversity in a tropical forest . Ecology 82 ( 4 ), 913 – 919 . Google Scholar CrossRef Search ADS 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 ( 3–4 ), 866 – 879 . Google Scholar CrossRef Search ADS Schwartz , G. , Pereira , P.C.G. , Siviero , M.A. , Pereira , J.F. , Ruschel , A.R. and Yared , J.A.G. 2017 Enrichment planting in logging gaps with Schizolobium parahyba var. amazonicum (Huber ex Ducke) Barneby: a financially profitable alternative for degraded tropical forests in the Amazon . For. Ecol. Manage. 390 , 166 – 172 . Google Scholar CrossRef Search ADS Sharma , L.N. , Grytnes , J.-A. , Måren , I.E. and Vetaas , O.R. 2016 Do composition and richness of woody plants vary between gaps and closed canopy patches in subtropical forests? J. Veg. Sci. 27 ( 6 ), 1129 – 1139 . Google Scholar CrossRef Search ADS Sipe , T. and Bazzaz , F. 1995 Gap partitioning among maples (Acer) in central New England: survival and growth . Ecology 76 ( 5 ), 1587 – 1602 . Google Scholar CrossRef Search ADS South , D.B. , Harris , S.W. , Barnett , J.P. , Hainds , M.J. and Gjerstad , D.H. 2005 Effect of container type and seedling size on survival and early height growth of Pinus palustris seedlings in Alabama, U.S.A . For. Ecol. Manage. 204 ( 2–3 ), 385 – 398 . Google Scholar CrossRef Search ADS Streit , K. , Wunder , J. and Brang , P. 2009 Slit-shaped gaps are a successful silvicultural technique to promote Picea abies regeneration in mountain forests of the Swiss Alps . For. Ecol. Manage. 257 ( 9 ), 1902 – 1909 . Google Scholar CrossRef Search ADS Therneau , T.M. 2018 coxme: Mixed Effects Cox Models. R package version 2.2–7. https://CRAN.R-project.org/package=coxme Vepakomma , U. , Kneeshaw , D. and Fortin , M.J. 2012 Spatial contiguity and continuity of canopy gaps in mixed wood boreal forests: persistence, expansion, shrinkage and displacement . J. Ecol 100 ( 5 ), 1257 – 1268 . Google Scholar CrossRef Search ADS Vilhar , U. , Roženbergar , D. , Simončič , P. and Diaci , J. 2015 Variation in irradiance, soil features and regeneration patterns in experimental forest canopy gaps . Ann. Forest Sci. 72 ( 2 ), 253 – 266 . Google Scholar CrossRef Search ADS Wang , G.L. and Liu , F. 2011 The influence of gap creation on the regeneration of Pinus tabuliformis planted forest and its role in the near-natural cultivation strategy for planted forest management . For. Ecol. Manage. 262 ( 3 ), 413 – 423 . Google Scholar CrossRef Search ADS Weber , T.A. , Hart , J.L. , Schweitzer , C.J. and Dey , D.C. 2014 Influence of gap-scale disturbance on developmental and successional pathways in Quercus-Pinus stands . For. Ecol. Manage. 331 , 60 – 70 . Google Scholar CrossRef Search ADS Webster , C.R. and Lorimer , C.G. 2005 Minimum opening sizes for canopy recruitment of midtolerant tree species: a retrospective approach . Ecol. Appl. 15 ( 4 ), 1245 – 1262 . Google Scholar CrossRef Search ADS Wright , E.F. , Coates , K.D. and Bartemucci , P. 1998 Regeneration from seed of six tree species in the interior cedar-hemlock forests of British Columbia as affected by substrate and canopy gap position . Can. J. For. Res. 28 ( 9 ), 1352 – 1364 . Google Scholar CrossRef Search ADS York , R.A. , Battles , J.J. , Eschtruth , A.K. and Schurr , F.G. 2011 Giant sequoia (Sequoiadendron giganteum) regeneration in experimental canopy gaps . Restor. Ecol. 19 ( 1 ), 14 – 23 . Google Scholar CrossRef Search ADS Zhu , J.J. , Liu , Z.G. , Wang , H.X. , Yan , Q.L. , Fang , H.Y. , Hu , L.L. , et al. 2008 Effects of site preparation on emergence and early establishment of Larix olgensis in montane regions of northeastern China . New Forests 36 , 247 – 260 . Google Scholar CrossRef Search ADS Zhu , J.J. , Lu , D.L. and Zhang , W.D. 2014 a Effects of gaps on regeneration of woody plants: a meta-analysis . J. Forestry Res. 25 ( 3 ), 501 – 510 . Google Scholar CrossRef Search ADS Zhu , J.J. , Wang , K. , Sun , Y.R. and Yan , Q.L. 2014 b Response of Pinus koraiensis seedling growth to different light conditions based on the assessment of photosynthesis in current and one-year-old needles . J. Forestry Res. 25 ( 1 ), 53 – 62 . Google Scholar CrossRef Search ADS © Institute of Chartered Foresters, 2018. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Forestry: An International Journal Of Forest ResearchOxford University Press

Published: Oct 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off